TIP5-CT-2006-031312 Page 1 of 289

URBAN TRACK Issued: 15/11/2010

D0602_M48_UITP_SP5.doc
DELIVERABLE D0602. FINAL CONSOLIDATED REPORT -
CHAPTER 5
Related Milestone
CONTRACT N 031312
PROJECT N FP6-31312
ACRONYM URBAN TRACK
TITLE Urban Rail Transport
PROJECT START DATE September 1, 2006
DURATION 48 months
Subproject SP6 SP6
Work Package WP6.2 Consolidation
Technical consolidation report on all validation
results (Chapter 5)
Written by Andr Van Leuven D2S
Frdric Le Corre / Laurent Illoz / Christine Charles /
Diego Sabato
ALSTOM /
POLIMI
Yves Amsler and Caroline Hoogendoorn UITP
Date of issue of this report 15 November 2010
PROJECT CO-ORDINATOR
Dynamics, Structures & Systems International D2S BE
PARTNERS
Socit des Transports Intercommunaux de Bruxelles STIB BE
Alstom Transport Systems ALSTOM FR
Bremen Strassenbahn AG BSAG DE
Composite Damping Materials CDM BE
Die Ingenieurswerkstatt DI DE
Institut fr Agrar- und Stadtkologische Projekte an
der Humboldt Universitt zu Berlin
ASP DE
Tecnologia e Investigacion Ferriaria INECO-TIFSA ES
Institut National de Recherche sur les Transports &
leur Scurit
INRETS FR
Institut National des Sciences Appliques de Lyon INSA-CNRS FR
Ferrocarriles Andaluces FA-DGT ES
Alfa Products & Technologies APT BE
Autre Porte Technique Global GLOBAL PH
Politecnico di Milano POLIMI IT
Rgie Autonome des Transports Parisiens RATP FR
Studiengesellschaft fr Unterirdische Verkehrsanlagen STUVA DE
Stellenbosch University SU ZA
Ferrocarril Metropolita de Barcelona TMB ES
Transport Technology Consult Karlsruhe TTK DE
Universit Catholique de Louvain UCL BE
Universiteit Hasselt UHASSELT BE
Project funded by the
European Community under
the
SIXTH FRAMEWORK
PROGRAMME
PRIORITY 6
Sustainable development,
global change & ecosystems International Association of Public Transport UITP BE
Union of European Railway Industries UNIFE BE
Verkehrsbetriebe Karlsruhe VBK DE
Fritsch Chiari & Partner FCP AT
Metro de Madrid MDM ES
Frateur de Pourcq FDP BE
Approved.doc
Quality checked & approved
by project co-ordinator Andr Van Leuven
TIP5-CT-2006-031312 Page 2 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
T AB L E OF CONT E NT S
5 Functional requirements for urban rail tracks........................................................................................... 6
5.1 Definition of Functional Track Requirements for Tram & Metro.................................................... 7
(a) Conclusions on Tramway Track form................................................................................... 7
(b) Conclusions on Metro Track form......................................................................................... 8
5.1.1 Modelling and simulations ............................................................................................................ 8
5.1.1.1. Tramway Track form............................................................................................................. 8
5.1.1.2. Metro Track form................................................................................................................... 8
5.1.2 Contact forces on tramway track................................................................................................... 9
5.1.2.1. Introduction............................................................................................................................ 9
5.1.2.2. Track irregularities and contact forces.................................................................................. 9
(a) Numerical model validation................................................................................................ 10
(b) Gauge irregularities vs. contact forces ................................................................................ 10
(c) Alignment irregularities vs. contact loads.......................................................................... 14
(d) Top irregularities vs. contact loads ..................................................................................... 17
(e) Cant irregularities vs. Contact loads................................................................................... 20
(f) Analysis................................................................................................................................. 23
5.1.2.3. Track alignment criterias optimization............................................................................... 24
(a) Principle of the study........................................................................................................... 24
(b) Preliminary studies of the dynamics calculations.............................................................. 25
(c) Applications to optimization of design criteria.................................................................. 27
5.1.3 Rail stress and deformation.......................................................................................................... 33
5.1.3.1. Introduction.......................................................................................................................... 33
5.1.3.2. Theoretical bases for the study............................................................................................ 34
(a) Working hypotheses ............................................................................................................ 34
(b) The stationary method......................................................................................................... 34
(c) Additional hypotheses......................................................................................................... 35
(d) Calculation methods ............................................................................................................ 36
5.1.3.3. Implementation under Castem............................................................................................ 39
(a) General .................................................................................................................................. 39
(b) Meshing................................................................................................................................. 39
(c) Specific procedures .............................................................................................................. 40
5.1.3.4. Study results ......................................................................................................................... 45
(a) Elastic shakedown example................................................................................................. 45
(b) Plastic shakedown example................................................................................................. 46
(c) Parametric study .................................................................................................................. 49
(d) Limits .................................................................................................................................... 56
(e) Future developments ........................................................................................................... 57
5.1.3.5. Conclusion............................................................................................................................ 58
5.1.3.6. Bibliographical references.................................................................................................... 59
TIP5-CT-2006-031312 Page 3 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.4 Measurements in Madrid............................................................................................................. 60
5.1.4.1. Introduction.......................................................................................................................... 60
5.1.4.2. Test Conditions..................................................................................................................... 60
(a) Test Vehicle........................................................................................................................... 60
(b) Track System Types.............................................................................................................. 61
5.1.4.3. Geometry Measurement ...................................................................................................... 63
(a) Interpretation of Readings................................................................................................... 63
(b) Statistic study of V1: [Puerta deBoadilla-Colonia Jardin].................................................. 65
(c) Statistic study OF V2: [Puerta de Boadilla-Colonia Jardin] ............................................... 78
5.1.4.4. Roughness Measurement..................................................................................................... 86
(a) The measurement trolley..................................................................................................... 86
(b) The measurements ............................................................................................................... 86
(c) Results................................................................................................................................... 90
(d) ISO 3095 ................................................................................................................................ 90
5.1.4.5. Pass-by Measurement: Noise and Vibration ...................................................................... 93
(a) Set-up .................................................................................................................................... 93
(b) Site S1 Classic + manta......................................................................................................... 93
(c) Results of noise measurements............................................................................................ 95
(d) Results of vibration measurements..................................................................................... 95
(e) Conclusion............................................................................................................................ 96
5.1.5 Metro track form........................................................................................................................... 97
5.1.5.1. The Mathematical Model ..................................................................................................... 97
(A) Train-Track Interaction Model ............................................................................................ 97
(B) Wheel-Rail Contact Model................................................................................................... 99
5.1.5.2. Metro Madrid ..................................................................................................................... 101
(a) Vehicle and Track Characteristics ..................................................................................... 101
5.1.5.3. Numerical Results .............................................................................................................. 105
5.1.5.4. References ........................................................................................................................... 109
5.1.6 General conclusion ..................................................................................................................... 110
5.1.6.1. Conclusions on Tramway Track form............................................................................... 110
(a) Track geometry................................................................................................................... 110
(b) Rail fatigue.......................................................................................................................... 110
(c) Track roughness ................................................................................................................. 110
(d) Vibration mitigation........................................................................................................... 110
5.1.6.2. Conclusions on Metro Track form..................................................................................... 111
TIP5-CT-2006-031312 Page 4 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.2 FONCTIONAL SPECIFICATIONS FOR TRACK INFRASTRUCTURE ..................................... 112
5.2.1 Sensitivity analysis on tramway track parameters by means of numerical simulation of train-
track dynamic interaction....................................................................................................................... 112
5.2.1.1. Introduction........................................................................................................................ 112
(a) Vehicle type ........................................................................................................................ 112
(b) Track type ........................................................................................................................... 113
(c) Rail irregularities................................................................................................................ 113
5.2.1.2. Description of the methodology........................................................................................ 113
(a) Calculation of the forces and displacements in the track ................................................ 114
(b) Calculation of stresses in the rail in MSC.Nastran........................................................... 115
5.2.1.3. Presentation of the models and their parameters............................................................. 116
(a) Presentation of the vehicle model ..................................................................................... 116
(b) Presentation of the track models ....................................................................................... 122
(c) Presentation of the track layout......................................................................................... 132
5.2.1.4. Results................................................................................................................................. 135
(a) Results out of Vi-Rail (time domain)................................................................................. 135
(b) Results out of Matlab ......................................................................................................... 137
(c) Results out of Nastran........................................................................................................ 142
5.2.1.5. Conclusion.......................................................................................................................... 154
5.2.2 Sensitivity analysis on metro track parameters by means of numerical simulation of train-
track dynamic interaction....................................................................................................................... 155
5.2.2.1. Overview............................................................................................................................. 155
5.2.2.2. Numerical simulation of train-track interaction............................................................... 155
(a) Vehicle model ..................................................................................................................... 157
(b) Track model ........................................................................................................................ 158
5.2.2.3. Simulations in tangent track.............................................................................................. 161
(a) Simulations performed ...................................................................................................... 161
(b) Track dynamic performance estimation: vibration mitigation........................................ 163
(c) Effect of track degradation................................................................................................. 164
5.2.2.4. Simulations in curve........................................................................................................... 167
(a) Simulations performed ...................................................................................................... 167
(b) Track dynamic performance estimation: track stability................................................... 167
5.2.2.5. Numerical-experimental comparison ............................................................................... 170
5.2.2.6. Concluding remarks........................................................................................................... 174
5.2.3 Light rail parameters based on national guidelines and recommendations........................... 175
5.2.3.1. Introduction: ....................................................................................................................... 175
5.2.3.2. Rules on the Alignment of Rail Systems........................................................................... 175
(a) General Objectives.............................................................................................................. 175
(b) Standard and Limit Values................................................................................................ 175
(c) Speeds ................................................................................................................................. 176
(d) Straight................................................................................................................................ 176
TIP5-CT-2006-031312 Page 5 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(e) Curve and Superelevation................................................................................................. 176
(f) Transition Curve and Superelevation Ramp.................................................................... 178
(g) Gradient and Change in Gradient..................................................................................... 180
(h) Deviations for Non-Standard Gauge Rail Systems .......................................................... 181
5.2.4 Applicability of rail standards to urban rail (tram& metro) ................................................... 182
5.2.5 Overview of a new embedded track standard under development in the US by
APTA/AREMA .......................................................................................................................... 182
5.2.5.1. Need for recommended practices ..................................................................................... 182
5.2.5.2. Five priorities...................................................................................................................... 182
5.2.5.3. Work plan ........................................................................................................................... 182
(a) Issues to be addressed........................................................................................................ 182
(b) Knowledge inventory......................................................................................................... 182
5.2.5.4. Actual progress - Embedded track.................................................................................... 183
5.2.5.5. Actual progress - non ballasted track................................................................................ 183
Appendix A5.2.1: Vibration velocity spectra with measured rail irregularities.......................................... 184
Appendix A5.2.2: Vibration velocity spectra with step function rail irregularity....................................... 225
Appendix A5.2.3 Applicability of Rail standards....................................................................................... 268
Appendix A5.2.4 Extracts from AREMA, chapter 12 part 8.................................................................... 269
TIP5-CT-2006-031312 Page 6 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5 FUNCTIONAL REQUIREMENTS FOR URBAN RAIL TRACKS
The following chapter 5 is presenting Sub-project 5 (SP5) targeting functional requirements for Urban
Rail tracks.
SP1 and SP2 see chapters 1 and 2 - focused respectively on new track construction and on track renewal
or maintenance. The purpose of SP5 performed by D2S, Alstom, TTK and Polimi was to identify where
the axes are for improvement and for further development of track components, construction
methodology and system design, and to set the basis for further evaluation of improvements and
development that resulted from SP1 and SP2 and which have been validated in SP3 (see chapter 3).
The sub-project 5 was made of two work packages:
WP5.1, presented in chapter 5.1, was targeting the definition of functional requirements and more
especially the determination and analysis of the forces applied to the track and the subsequent
stress and displacement in rail. Another objective was to assess the influence of the track condition
(geometrical and mechanical) on the vehicle dynamic and the subsequent impact on stress in rail.
WP5.2, presented in chapter 5.2, concentrated on the definition of functional specifications
regarding Tram, Light Rail and metro tracks. This objective could be achieved through the following
approach:
by defining the significant parameters for identifying the level of duty conditions (loads): track
characteristics and conditions, vehicle characteristics and conditions, traffic type and density.
by identifying the range of duty conditions that exists on segments of the track infrastructure of
participating networks taking into account the influence of vehicle characteristics and condition
and traffic density.
by selecting, through consultation with network operators, those duty conditions and track
parameters that could provide the most marked decreases in LCC through the implementation of
improved products and processes. In addition to the use of improved products in track renewals,
the objective was also to increase the residual life of existing track through the use of better
maintenance practices.
Updated functional specifications for a track infrastructure in a new network were made available, also
based upon:
sensitivity analysis and parametric studies which evaluate the influence of track parameters (such as
gauge, gauge widening, rail type, rail support stiffness, gauge tolerance, ) on track stability, on
noise & vibrations, on construction costs and on LCC in general;
the findings in ongoing EC research projects such as CORRUGATION, INMAR, TURNOUTS,
SPURT, MODTRAIN, WIDEM, SILENCE, QCITY.
At the end of the sub-project, it has also been decided to carry out a thorough analysis of existing railway
standards: since track alignment, track irregularities are defined in railway standards primarily from
mainlines networks, the purpose was to review these standards and to point out their relevance
regarding urban rail systems, as well as the needs for adaptation in order to better cope with urban rail
TIP5-CT-2006-031312 Page 7 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
requirements. This analysis is presented in separate EXCEL files and the outcomes were used as an input
for the works of the Urban Rail Platform set up by UITP and UNIFE to develop recommendations for
technical harmonisation of urban rail systems throughout Europe.
5.1 DEFINITION OF FUNCTIONAL TRACK REQUIREMENTS FOR TRAM&METRO
Within WP5.1, the work was developed through models and simulations on one hand and site
measurement on the other hand, for both tram and metro tracks. Vehicle dynamic simulations were
carried out based on multi-body method in order to have a complete view of the track response, to assess
the influence of passing vehicles and to evaluate the sensitivity of the wheel rail interaction for various
track systems. The subject of rail stress has been discussed in terms of fatigue behaviour under repeated
cyclic loading. The dynamic forces were generated by railway vehicles and also by track irregularities.
Fracture mechanics was applied to assess the behaviour of the rail. A new method for simulating a high
number of passes was implemented.
(a) CONCLUSIONS ONTRAMWAY TRACK FORM
A wide range of subjects was dealt with in the present document, all of them related to the forces applied
to the track and the resulting stresses in rail, and leading to conclusions on track geometry, track
roughness and vibration mitigation.
Complete models including track and vehicles were setup allowing for simulations taking into account
the actual conditions of the track.
(i) Track geometry
The track geometry was studied in relation with its mechanical response and the resulting forces and
stresses undergone. Following the measurements in Madrid of track for gauge, vertical alignment, and
cant and horizontal alignment, an approach based on 95% compliance can be applied to conclude that the
track V1 of the ML-3 line is compliant to the EN 13231-1 standard acceptance tolerances for these track
parameters.
The results of the track geometry study can be proposed as a contribution for the development of urban
track standards.
(ii) Rail fatigue
The rail fatigue was studied with two practical purposes: to be able to make a commitment on the lifetime
of the rails installed and to optimise the rail replacement frequencies for the track maintenance.
The development of that study was the opportunity to implement the stationary method, which enables
to undertake fast simulation of a high number of cycles.
(iii) Track roughness
The track roughness measured on Madrid network is particularly good in comparison with what could
be found on some other networks.
The reason for this good track roughness is somewhat difficult to find even after analysis of measurement
collected.
TIP5-CT-2006-031312 Page 8 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
It is difficult to draw a conclusion at the moment.
(iv) Vibration mitigation
The approach pertaining to vibration mitigation has to consider the whole vibration transmission path
and to be adapted to the vehicle that will be operated.
And the vibration performance of a transportation system must be defined in relation to its compliance
with a level of vibration measured at building locations and not by empirical rules based on outdated
data.
The measurements made on many other sites for several tram designs and manufacturers are correlating
what could be observed in Madrid:
A very high performance system (floating slab) is rarely required,
A continuously supported rail system is sufficient in many cases.
(b) CONCLUSIONS ON METROTRACK FORM
The study of the Madrid metro track form based on the REMS system was finalised following the
calibration and adjustment phases carried out during laboratory testing.
5.1.1 Modelling and simulations
The determination and analysis of track forces and rail stress data addresses two types of track forms:
tramway and metro.
5.1.1.1. Tramway Track form
As a basis, a typical description of the tramway track form was provided. Then modelling and simulation
works of selected tramway track forms were carried out.
The linear dynamic response, under a permanent regime, of a multi-layer half space subjected to a mobile
load applied to its free surface, was studied. The purpose was to validate the quasi-static approach, as the
influence of speed is not major in urban railway context.
Following this validation, two modelling approaches were investigated:
Semi Analytical
Numerical
The results were quite similar between the two approaches with an advantage for the semi-analytical as
the simulation time is much reduced.
5.1.1.2. Metro Track form
The first step in the study of the metro track form consisted in setting-up the mathematical model and in
particular the train-track interaction model and the wheel-rail contact model.
The metro of Madrid is the reference of the current study. Vehicle and track characteristics were input in
the models.
Finally the numerical results were output. Displacements and forces on track were computed.
TIP5-CT-2006-031312 Page 9 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.2 Contact forces on tramway track
5.1.2.1. Introduction
The rails installed on a track typically undergo forces of different nature: static, quasi-static (e.g. in
curves), and dynamic.
The current section will focus on quasi-static and dynamic forces as they are in most cases at the origin of
rail degradation.
The source of these forces lies, on one hand, in the dynamic behaviour of the rolling stock and, on the
other hand, in the track and rail conditions.
The phenomena occurring within the wheel-rail contact area because of geometrical imperfection of
wheel or rail and because of resonance phenomena due to the various stiffnesses involved in the wheel
suspension and in the rail supporting.
One of the roles of the track is to guide the rolling stock so that it follows a determined path, which is
necessarily made of straight alignment, transitions and curves. These two latter types of section induce a
dynamic behaviour from the rolling stock.
The study developed below will focus on the influence of track alignment on train dynamics with the
objective of determining the technical-economical optimum for track geometry and tolerances and for
track alignment.
5.1.2.2. Track irregularities and contact forces
The objective of this study is to analyse the impact of the track irregularities on the dynamic behaviour of
the rolling stock (tramway). Four track geometry parameters are particularly under interest: The gauge,
the alignment (or horizontal level), the top (or vertical level), and the cant.
Track geometry irregularities used for the entire analysis are extracted from in-field measurements. All
measured irregularities are then numerically increased or decreased. The variation of these four
parameters is chosen in order to check whether EN 13231-1 (new construction) is applicable to urban
transportation such as tramways.
All parameters are analysed individually, in other terms no combination of defects has been studied.
For each parameter and for each numerical configuration, safety criteria, passenger comfort and
wheel/rail contacts loads are assessed and analysed with respect to the amplitude of track geometry
irregularities.
The results of this study allow confirming the applicability of EN 13231-1 tolerances used for the track
works of a new line in an urban context and with the Citadis tramway.
TIP5-CT-2006-031312 Page 10 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(a) NUMERICAL MODEL VALIDATION
The numerical model used for the study has been correlated from measurements done on a French
network. The validation process is the same as the one described in section 5.1.2.3.b.
(b) GAUGE IRREGULARITIES VS. CONTACT FORCES
The initial gauge defects extracted from measurements are displayed in the picture below.
Figure 5.1.1 Measured gauge irregularities
This defect is then amplified and/or lowered as shown in the picture below:
Figure 5.1.2: Amplification/Reduction of gauge irregularities amplitude
EN 13231 -1
TIP5-CT-2006-031312 Page 11 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
As can be seen in the above picture, the case where initial measurements are multiplied by 2 is, in certain
areas above the limit given in EN 13231-1. Other cases are mainly compliant with gauge irregularities
tolerances of EN 13231-1.
As previously mentioned, the followings items are further analysed: Safety parameters, Lateral and
Vertical dynamic loads at wheel/rail contact, maximum lateral and vertical accelerations
(i) Safety Parameters:
Safety parameters are derailment parameter (Y/Q ; Y, Q respectively the transversal ans vertical load),
Prudhomme criteria ( ( ) ) 3 10 (
0 lim 2
P Y
m
+ = o E ) , dynamic wheel unloading (dQ/Q ; dQ the dynamic
vertical load). The safety parameters are checked for each dynamic calculation with respect to the criteria
given in the table below.
SAFETY PARAMETERS CRITERIA
Y/Q Y/Q 0,94
CITADIS 402
Prudhomme
= 1
P = axle Static Load
(Sy Essieu1)limit = 33.79 KN
(Sy Essieu2)limit = 33.33 KN
(Sy Essieu 3)limit = 31.81 KN
(Sy Essieu 4)limit = 31.78 KN
(Sy Essieu 5)limit = 32.95 KN
(Sy Essieu 6)limit = 32.89 KN
(Sy Essieu 7)limit = 33.06 KN
(Sy Essieu 8)limit = 33.54 KN
dQ/Q dQ/Q 0,6
Table 5.1.1 Safety Parameters - Criteria
As can be seen in the table 5.1.2, the safety parameters are in any case under the limit given above.
Safety Parameters: Maximum Values calculated
Gauge
Y/Q Max dQ/Q
Prudhomme
S(y) in KN
Criteria
0.04 0.02 0.46 OK
0.03 0.02 0.35 OK
0.05 0.02 0.54 OK
0.05 0.03 0.68 OK
0.11 0.05 1.34 OK
Table 5.1.2 Safety Parameters - Maximum Values
Increasing
Gauge
TIP5-CT-2006-031312 Page 12 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii) Dynamic loads on the wheel/rail contact:
The picture below displays the influence of track gauge irregularities on transversal loads.
Figure 5.1.3: Effort Y - Influence of gauge irregularities
As can be seen, only the last case (i.e. initial times 2, this case corresponds to the case where gauge is
above the limits specified in EN 13231-1) gives rise to high levels of lateral loads.
The picture below displays the influence of track gauge irregularities on vertical loads.
Figure 5.1.4: Loads Q Gauge irregularities
TIP5-CT-2006-031312 Page 13 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
As can be seen, there is no influence of track gauge irregularities, in the variation range studied, on the
vertical loads.
(iii) Lateral and vertical accelerations:
Only the results calculated at axle levels are shown because the values are more significant.
The picture below displays the influence of track gauge irregularities on axle lateral rms acceleration.
Figure 5.1.5: Lateral rms acceleration - Influence of gauge irregularities
As can be seen, only the last case (i.e. initial times 2, this case corresponds to the case where gauge is
above the limits specified in EN 13231-1) gives rise to high levels of lateral accelerations.
The results obtained for the axle vertical rms acceleration show that, as per vertical load, the gauge
irregularities have almost no influence on this parameter.
The same analysis is performed for all other parameters; only the main results are given in following
sections.
TIP5-CT-2006-031312 Page 14 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(c) ALIGNMENT IRREGULARITIES VS. CONTACT LOADS
The initial alignment irregularities extracted from measurements are displayed in the picture below
(black curve). Other curves correspond to reduction (green curve) or amplification (other curves) of the
initial alignment signal.
Figure 5.1.6: Amplification / Reduction of alignment irregularities
As shown in the picture above, the cases where initial measurements are multiplied by 2.6 and 2.9 are, in
certain areas above the limit given in EN 13231-1. Other cases are fully compliant with alignment
irregularities tolerances of EN 13231-1.
As per gauge irregularities, followings items are further analysed: safety parameters, lateral and vertical
dynamic loads at wheel/rail contact and maximum lateral and vertical accelerations.
(i) Safety Parameters:
The table 5.1.3 below shows the main results for safety parameters:
Safety Parameters: Maximum Values calculated
Alignment
Y/Q Max dQ/Q
Prudhomme
S(y) en KN
Criteria
0.16 0.1 5.08 OK
0.09 0.09 3.53 OK
0.34 0.15 5.99 OK
0.45 0.23 10.67 OK
0.51 0.27 12.66 OK
Table 5.1.3: Safety parameters - Maximum values calculated
All Safety parameters are under the criteria. So no alignment configuration tested is critical.
Increasing
Alignment
EN 13231 -1
TIP5-CT-2006-031312 Page 15 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii)Dynamic loads on the wheel/rail contact:
c.ii-1 Lateral Loads:
The picture below displays the influence of track alignment irregularities on lateral loads.
Figure 5.1.7: Mean values of Lateral loads Horizontal alignment
In the above picture, it can be noted that, while alignment irregularities are under the levels specified in
EN 13231-1 (i.e. [-5;+5]mm on a 10 m chord), the values of lateral loads are low and have almost the same
value. These cases correspond to plain bars. As the EN13231-1 threshold is overcome, the lateral loads
amplitude increase. These cases correspond to bars with stripes.
c.ii-2 Vertical Loads:
It was found that there is no influence of track alignment irregularities (in the range under analysis) on
vertical loads.
TIP5-CT-2006-031312 Page 16 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(iii) Vertical and lateral acceleration:
Only the results at axle levels are shown.
c.iii-1 Lateral Acceleration:
The picture below displays the influence of track alignment irregularities on axles rms transverse
acceleration.
Figure 5.1.8: RMS Transverve acceleration per axle
In the above picture, it can be noted that, while alignment irregularities are under the levels specified in
EN 13231-1 (i.e. [-5;+5]mm on a 10 m chord), the values of lateral acceleration are low and remain
equivalent. These cases correspond to plain bars. As the EN 13231-1 threshold is overcome, the lateral
acceleration increases. These cases correspond to bars with stripes.
TIP5-CT-2006-031312 Page 17 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(d) TOP IRREGULARITIES VS. CONTACT LOADS
The initial top irregularities extracted from measurements are displayed in the picture below (black
curve). Other curves correspond to reduction (green curves) or amplification (other curves) of the initial
top signal.
Figure 5.1.9: Track irregularuitites
As shown in the picture above, the case where an initial measurement is multiplied by 3.5 is, in certain
areas above the limit given in EN 13231-1. Other cases are mainly compliant with top irregularities
tolerances of EN 13231-1.
(i) Safety Parameters:
The table below shows the main results with respect to safety parameters:
Safety Parameters: Maximum Values calculated
Top Level
Y/Q Max dQ/Q S(y) en KN Criteria
0.02 0.07 0.18 OK
0.02 0.03 0.18 OK
0.02 0.04 0.18 OK
0.02 0.23 0.46 OK
0.02 0.25 0.50 OK
Table 5.1.4: Safety Parameters: Maximum Values calculated
All Safety parameters are under the criteria. So no top-level configuration tested is critical.
EN 13231 -1
Increasing
Top Level
TIP5-CT-2006-031312 Page 18 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii) Dynamic loads on the wheel/rail contact:
d.ii-1 Lateral Loads:
Lateral loads do not vary significantly with top-level amplification. Thus, in the variation range studied,
no significant influence has been demonstrated between top level and lateral loads.
d.ii-2 Vertical Loads:
Vertical loads do not vary significantly with top-level amplification. Thus, in the variation range studied,
no significant influence between top level and vertical loads has been demonstrated.
TIP5-CT-2006-031312 Page 19 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(iii) Vertical and Lateral acceleration:
d.iii-1 Lateral acceleration:
In any configuration tested (see picture below), lateral accelerations have very low values, which means
that the influence of top-level irregularities on lateral RMS acceleration is not significant.
Figure 5.1.10: Axle vertical acceleration
d.iii-2 Vertical acceleration:
Vertical accelerations vary when the amplitude of the top-level irregularity is close or above the limit
mentioned in EN 13231-1 ([-6; +6] mm). Nonetheless, the values calculated remain low as can be seen in
the following picture.
Figure 5.1.11: Axle vertical acceleration
TIP5-CT-2006-031312 Page 20 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(e) CANT IRREGULARITIES VS. CONTACT LOADS
The initial cant irregularities extracted from measurements are displayed in the picture below (black
curve). Other curves correspond to reduction (green curves) or amplification (red or orange curves) of the
initial cant signal.
Figure 5.1.12: Amplification / Reduction of cant irregularities
In the picture above, the cases where initial measurements are multiplied by 3.8 and 4.5 are, in certain
areas above the limit given in EN 13231-1. Other cases are mainly compliant with cant irregularities
tolerances of EN 13231-1.
(i) Safety Parameters:
The table below shows the main results for parametric study on cant irregularities with respect to safety
parameters:
Safety Parameters: Maximum Values calculated
Cant irregularities
Y/Q Max dQ/Q S(y) en KN Criteria
0.02 0.04 0.63 OK
0.03 0.03 0.40 OK
0.04 0.10 1.66 OK
0.04 0.16 2.41 OK
0.04 0.18 2.38 OK
Table 5.1.5: Safety Parameters - Cant irregularities
All safety parameters are under the criteria. So no can irregularities configuration tested is critical.
EN 13231 -1
Increasing
Cant irreg.
TIP5-CT-2006-031312 Page 21 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii) Dynamic loads on the wheel/rail contact:
e.ii-1 Lateral Loads:
Lateral loads only vary significantly when cant irregularities correspond to higher levels than those given
in EN 13231-1 (see picture below).
Figure 5.1.13: Mean value of lateral loads
The case corresponding to the where the initial cant irregularities amplitude is multiplied by 2.6 is
satisfactory with regards to lateral loads. This latter case corresponds to a cant irregularity signal
equivalent to EN 13231-1.
e.ii-2 Vertical Loads:
Lateral loads do not vary with cant irregularities (See Picture below).
Figure 5.1.14: Vertical loads - Influence of of cant irregularities
TIP5-CT-2006-031312 Page 22 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(iii) Vertical and Lateral acceleration:
e.iii-1 Lateral acceleration:
As can be seen in picture below, only the last two cases (i.e. initial times 3.8 and initial times 4.5, this case
corresponds to the case where Cant irregularities is above the limits specified in EN 13231-1) gives rise to
higher levels of lateral accelerations.
Figure 5.1.15: Lateral acceleration - Influence of cant irregularities
e.iii-2 Vertical acceleration:
The values of vertical accelerations obtained for this parametric study are very low. Moreover, no
significant variation was observed with varying cant irregularities. Thus, no influence of cant
irregularities on axle vertical acceleration is found in the studied variation range.
TIP5-CT-2006-031312 Page 23 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(f) ANALYSIS
Four geometrical parameters have been studied: Gauge, Alignment, Top level, Cant. Those parameters
have been studied in terms of irregularities, in other terms deviation from theoretical values.
The objective of this study was to check whether the EN 13231-1 (for new track construction) could be
applied to typical tramway applications. In order to answer this question, four parametric studies have
been performed, one for each geometrical parameters. Basically, the parametric studies consisted in
amplifying or reducing the levels of irregularities and simulating the influence of such an operation on
safety parameters, vertical and lateral loads, vertical and lateral accelerations.
These four parametric studies enable to conclude that for:
Safety parameters: the safety parameters are fully and largely compliant with the safety criteria. In other term,
the values given in EN 13231-1 (for new track construction) present no safety risk.
Vertical and Lateral Loads: as long as track irregularities are below EN 13231-1, no significant influence of those
defects has been observed on vertical and lateral loads.
Vertical and Lateral Accelerations: since track irregularities are below EN 13231-1, no adverse influence of those
defects has been calculated on axle vertical and lateral accelerations.
In the light of those observations/results, it can be concluded that no adverse situation is expected in the
case EN 13231-1 (new construction) is applied. Thus, EN 13231-1 (new construction) could be applied to
tramway application.
TIP5-CT-2006-031312 Page 24 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.2.3. Track alignment criterias optimization
The objective of this study is to analyse the impact of the track alignment criteria on the wheel rail contact
and forces
(a) PRINCIPLE OF THE STUDY
(i) Model validation
The validation of the model used for this study was conducted on the basis of tests conducted on a
French network.
(ii) Parametric study
a.ii-1 Specification of the parametric study
The parametric study allows testing a number of configurations of track design by varying the
different components of the track one after another. For each configuration, a number of parameters
are evaluated to valid whether or not a configuration.
The following configurations will be a parametric study.
Studies of track alignment criteria will focus on following cases:
Length of Clothoid ( = Transition) (for Radius of curvature R = 25, 50, 100, 200, 300, 400, 500 m)
Length of straight alignment between two clothods in case of reverse curves (for Radius of curvature
R= 25, 50, 100, 200, 300, 400, 500 m),
Overlapping (combination of Vertical and Horizontal Curves),
a.ii-2 Evaluation parameters
The security parameters are 3 and will be systematically checked for each dynamic calculation:
The derailment indicator Y/Q: 0.95
The sum wheelset calculated on each wheelset ( ) ) 3 10 (
0 lim 2
P Y
m
+ = o E
The wheel unloading behaviour dQ/Q 0.6
The classic parameters to evaluate are accelerations in the different carbody and are used to calculate the
Jerk.
The vertical and lateral loads will be evaluated at the contact for all wheels from the rolling stock (not
presented in this document).
a.ii-3 Validation of all tested configurations
All tested configurations will be validated if all security parameters are respected, then the comfort
criteria and effort to the contact patch can be optimized.
TIP5-CT-2006-031312 Page 25 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(b) PRELIMINARY STUDIES OF THE DYNAMICS CALCULATIONS
(i) General validation of the model on vampire software
The parametric study is performed using VAMPIRE dynamic simulations software. To ensure the
validity of the results, a validation of the model is preferable. This is done on the results of the
measurement campaign of the network. An initial check was conducted on the basis of accelerometer
placed on the vehicle (on the floor on the carbody M2). The comparison between numerical predictions
and simulation results show a very good correlation of the model.
Figure 5.1.16: Comparison measurement / simulation acceleration in Carbody M2
Measurements of track defects measured on the site of The network (gauge, alignment, top, cant) are
located every 25 cm and are implemented in the dynamic simulation program to simulate the dynamic
behaviour of the tramway as realistic as possible.
Results of amplitudes and signal dynamics form remain satisfactory as seen in the graph above.
TIP5-CT-2006-031312 Page 26 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii) Length of curvature taken in dynamics calculations
b.ii-1 Objective
To determine the minimum length of curvature for radius R = 25 to 500 m from dynamic calculations in
order to avoid any interaction on others parameters for example when we study the influence of length of
the clothoid. These minimum length of curvature will be kept in all future calculations.
b.ii-2 Summary of minimum curvature length for all radius
Below trends representing the minimum arc length for each radius of curvature for Citadis 302_402.
Tendance des Larc "optimale" en fonction du Rayon de l'arc de cercle
-Citadis 302_402-
50
60
70
80
90
100
110
120
130
140
0 100 200 300 400 500 600
Rayon de l'arc de cercle (m)
L
a
r
c
O
p
t
i
m
a
l
e
(
m
)
Tendance Larc Optimale pour chaque Rayon_ Citadis 402 Duba
Tendance Larc Optimale pour chaque Rayon_ Citadis 302 Montpellier
Radius of curvature (m) Curvature length 302_402 taken in calculations (m)
25 63
50 68
100 79
200 95
300 112
400 122
500 130
Table 5.1.6: Radius vs Curvature
The length of curvature according the radius will be taken into account in the construction of track file for
dynamic calculations in order to study the influence of the clothoid length, the length of alignment
between reverse curve, overlapping (without the impact of the circular curve length).
TIP5-CT-2006-031312 Page 27 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(c) APPLICATIONS TOOPTIMIZATION OF DESIGN CRITERIA
(i) Minimum length of clothoid (transition) in curve without cant
c.i-1 Objective
Determine limpact of the lengths of clothoid on the behaviour of rolling stock for radius of curvature
from R = 25 m to R = 500m. These simulations will also verify the behaviour of RS in the minimum values
calculated according to the specification of track alignment criteria for Citadis 302_402 and will help to
validate or optimize the calculation method following.
The maximum speed (in circular curve without cant) is calculated with respect to the maximum value of
lateral acceleration of 0.68 m/s
2
following the formula. This formula reflects the maximum allowable
lateral acceleration of 0.68 m/s
2
.in curve of constant radius.
R V . 68 , 0 6 , 3 = WhereV inkm/handR inm
In order to determine the optimum length of each clothoid (in circular curve without cant), the
following formulae are used in conjunction with the required speed and radius of curve (s).
.
R
v
L
3
5 , 2 = (v in m/s, R in m, J in m/s
3
)
The results of this formula will be compared with those of the dynamic calculations in order to validate
or optimize the length of clothoid.
c.i-2 Safety assessment in clothoid
All tested configurations respect the safety criteria (derailment, sum of lateral effort calculated on each
wheelset, wheel unloading) with different lengths of clothoid considered in dynamic calculations.
The derailment indicator Y/Q: 0.94
The sum of lateral effort calculated on each wheelset ( ) ) 3 10 (
0 lim 2
P Y
m
+ = o E
The wheel unloading behaviour dQ/Q 0.6
We can see the degradation risk increase for values inferior to calculated values by the formulae below
R
v
L
K
3
5 , 2 =
All these configurations (except the length of clothoid for R25 and R50 with 4m) respect safety criteria
and can be studied of point of view comfort passengers.
TIP5-CT-2006-031312 Page 28 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
c.i-3 Graph of the minimum clothoid length versus radius of curvature (without cant)
Longueur de clothode minimale vs Rayon de courbure
R=25 500m (sans dvers)
6
11
16
21
26
31
25 50 75 100 125 150 175 200 225 250 275 300 325 350 375 400 425 450 475 500
Rayon de courbure (m)
L
o
n
g
u
e
u
r
d
e
l
a
c
l
o
t
h
o

d
e
(
m
)
Longueur minimale calcule des simulations
dynamiques (valeurs optimises)
Longueur minimale calcule suivant
l'expression analytique
Figure 5.1.17: Minimum clothoid length vs Radius
This chart can be used to define the lengths of clothoid optimized for a certain radius of curvature. We
respect the maximum Jerk of 0.4 m/s
3
and the maximum lateral acceleration of 0.68 m/s
2
.
c.i-4 Conclusion: Length of the minimum clothoid (without cant), optimisation
RADIUS OF CURVATURE Minimum length of clothoid
(dynamic simulation)
Minimum length of clothoid
(according to the specification)
25 7 12 (7 exceptionally)
50 9.5 12 (10 exceptionally)
100 13.5 14
200 19 20
300 23 24.3
400 27 28
500 30 31.34
Table 5.1.7: Radius vs minimum length of clothoid
It is found that the calculated clothoid length following the formulation
R
V
L
3
5 , 2 = is higher than those
derived from dynamic simulations.
Optimized
TIP5-CT-2006-031312 Page 29 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii) Length of straight alignement between 2 clothoids in case of reverses curves
c.ii-1 Objective
o Check and optimize the value of 12 m imposed by the criteria of the specification and issued of
railway practices (EN 13803-1 et 2 design of track with abrupt changes of curvature) LAD mini = 0,2.V
with V the speed (km/h).
o Check and validate the value of 0 m (inflection point) in relation to the behaviour of Citadis in these
configurations.
1 00 2 00 3 00
- 10
-5
0
5
10 1/ km
1 00 2 00 3 00
- 10
-5
0
5
10 1/ km
1 00 2 00 3 00
- 10
-5
0
5
10 1/ km
The minimum lengths of clothoid (transition curves), which were described and validated in the previous
chapter for maximum speeds in curves without cant, are retained in this study. The dynamic calculations
were made in taking into account straights of 0, 3, 6, 9 and 12m between reverses curves of radius R =25
to 500 m. In all configurations the acceleration for calculation of the jerk was evaluated in the area
marked in brown to know the influence of the alignment between reverses curves (S-CURVE).
c.ii-2 Conclusion: Length of the minimum alignment between s-curves
Radius of curvature (in m) Length of minimal alignment
between s-curves (DYNAMIC
SIMULATION)
Length of minimal alignment
between s-curves
(SPECIFICATION CITADIS)
25 3 12
50 3 12
100 3 12
200 3 12
300 3 12
400 3 12
500 3 12
Table 5.1.8: Minimum length of alignnment in s-curve
The minimum length of alignment between S-curves will be 3 m whatever the studied radius of
curvature: respect of the maximum acceleration of 0.68 m/s
2
and a maximum Jerk of 0.4 m/s
3
Alignment is 6m
Alignment is 15m
Alignment is 0m
Optimized
TIP5-CT-2006-031312 Page 30 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(iii) Minimum length of clothoid for curves with cant
c.iii-1 Objective
Determine the minimum length of clothoid for radius of curvature with CANT from dynamic
simulations in order to verify if the formulation used until today in our design criteria is correct to
calculate the limit value of the rate of change of cant deficiency (dI/dt)lim in mm/s. This value
(dI/dt)lim is a parameter which is not a normalized data for tramways.
The speed used in dynamic simulations is the maximum speed using the equilibrium cant dth which is
the sum of cant with the cant deficiency. We respect a maximum acceptable lateral acceleration of 0.68
m/s2
Vmax = (R* dth/11.8) x0.5 (with equilibrium cant dth in mm and R in m, V in km/h)
and dth (equilibrium cant in mm) = applied cant + cant deficiency.
To dimension the length of clothoid in curve with cant, we use the 3 formulations according to the
standard PrEN 13803-1:
The minimum length of the transition curve (clothoid) will be the largest of these 3 formulations in order
to ensure the 3 limits defined above (respect of 3 conditions of the standard PrEN 13803-1).
c.iii-2 Used Methodology
The main objective of this study will be to determine a limit of Rate of change of cant deficiency
(dI/dt)lim from this methodology by studying different radius of curvature, cant and maximum speed
associated.
The dynamic simulations will verify and validate the procedure described below in defining a limit value
for the Rate of change of cant deficiency: dI/dt (using as criterion the maximum jerk to 0.4 m/s
3
as in the
case of curve without cant, the jerk defines the comfort felt by passengers). From the dynamics
| | m
dt
dD
D
V
L
D
1
lim
6 . 3

|
.
|

\
|
A >
| | m
ds
dD
D L
D
1
lim

|
.
|

\
|
A >
| | m
dt
dI
I
V
L
K
1
lim
6 . 3

|
.
|

\
|
A >
Formulae with the cant gradient: dD/ds
(dD/ds)
lim
= 3mm/m
Formulae with the Rate of change of cant: dD/dt
(dD/dt)
lim
=* V/ L
D
= 50 mm/s
Formulae with the Rate of change of cant deficiency: dI/dt
(dI/dt)
lim
= numerical data not normalized for
tramways ?
TIP5-CT-2006-031312 Page 31 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
calculations and according to EN 12299, the maximum jerk in clothoid will be calculated as in the case of
the study of curves without cant (see in the first chapter of this document).
Until today, the limit of the Rate of change of cant deficiency used was 0.61 mm/s, data provided by
EGIS-Rail recommendations. We will check this information in insuring that we respect a maximum jerk
of 0.4 m/s
3
Results of dynamic calculations will be compared with analytical calculations to verify the method used
and determine / validate a limit of the Rate of change of cant deficiency (dI/dt)lim.
c.iii-3 Conclusion: Accepted limit value of rate of change of cant deficiency used
(dI/dt)lim
Taking as limit value a rate of change of cant deficiency (dI/dt)lim = 61 mm/s, we respect a maximum
lateral acceleration of 0.68 m/s
2
in the track plan and a maximum jerk of 0.4 m/s
3
in the track plan
(conception criteria for the track).
The two methods (analytical and numerical) give identical values at the track plan, which means that the
methodologies used to calculate the Jerk by these methods are correct.
For the jerk calculated to the floor, there is a slight difference because the analytical method is simplified.
Its clear that the dynamic simulation results are more accurate to the floor because we take into account
the realistic dynamic behaviour of the train when it passes along the curve.
We therefore validated all analytical and dynamics calculations and have shown that the methods were
equivalent. The value of 0.61 mm/s for the rate of change of cant deficiency is justified.
See details of calculations below.
c.iii-4 Details of calculation of limit value of rate of change of cant deficiency (dI/dt)lim
Below the cases studied by the analytical method (developed in using PrEN 13803-1) and the dynamic
method on VAMPIRE (calculations of acceleration and jerk is then calculated in according to PrEN
12299), the values demonstrate the equivalence between these 2 methods and give us the limit value of
Rate of change of cant deficiency (dI/dt)lim ensuring a maximum Jerk = aq de 0.4 m/s
3
to the track
plan.
TIP5-CT-2006-031312 Page 32 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Radius of
curvature and
length of clothoid
associated
Applied Cant
and Deficiency
Maxim
um
speed
Analytical or
simulations
calculations
Lateral
acceleration to
track plan aq and
in carbody ai
Comparison between analytical or
simulations calculations
Lateral maximum Jerk calculated in
m/s
3
aq: Maximum Jerk to track plan
ai: Maximum Jerk in carbody
Cases where Lk is greater
to the two others
expressions, by default we
have taken 61 mm/s to
verify the maximum Jerk
associated (value taken
today by ALSTOM to
define the minimum length
of clothoid in curve with
cant)
Calculs (1) Simulations (2) R
(m)
Lk
(m)
D
(mm)
I
(mm)
V
(Km/h)
aq
(m/s
2
)
ai
(m/s
2
)
aq ai aq ai
Limit value of Rate of
change of cant deficiency
(dI/dt)lim in mm/s
100 15.35 20 104 32.4 0.68 0.80 0.399 0.467 0.392 0.429 61
100 17 20 104 32.4 0.68 0.80 0.360 0.421 0.361 0.399 55
100 18.73 20 104 32.4 0.68 0.80 0.327 0.383 0.329 0.368 50
100 19.71 40 104 34.9 0.68 0.80 0.335 0.392 0.336 0.400 51.2
100 21.03 60 104 37.3 0.68 0.80 0.335 0.392 0.335 0.422 51.2
300 31.68 20 104 56.1 0.68 0.80 0.335 0.392 0.336 0.393 51.2
300 34.14 40 104 60.5 0.68 0.80 0.335 0.392 0.335 0.408 51.2
300 36.43 60 104 64.6 0.68 0.80 0.335 0.392 0.334 0.421 51.2
500 39.22 10 104 69.5 0.68 0.80 0.335 0.392 0.335 0.386 51.2
Table 5.1.9: Detail of calculations of limit value of rate of change of cant deficiency
| | m
dt
dI
I
V
L
K
1
lim
6 . 3

|
.
|

\
|
A >
TIP5-CT-2006-031312 Page 33 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.3 Rail stress and deformation
5.1.3.1. Introduction
A rail is subject to strains and deformations due to the sheer number of trains passing over it. Each time a
wheel passes over may be considered as a load cycle. The accumulation passing wheels of a rolling stock
are exerting a repeated loading on that rail. And as a response to this undergone cyclic loading it has
been observed that the rail material is reaching an asymptotic regime, which can be:
Pure elasticity,
Elastic shakedown,
Plastic shakedown,
Ratchetting.
Identifying which one of the above regimes is reached, provides an indication about the subsequent
consequences due to the numerous passes of trains and constitutes a first step in determining the
expected life duration of the rail.
Several parameters are determining the trend the material will follow and the nature of the asymptotic
regime. So we will firstly present the calculus method known as stationary method as well as the
parameters identified as relevant ones. Then the implementation of the stationary method with the FEM
software Castem will be detailed. Finally, the results of the parametric study will be given.
They will lead into a conclusion concerning the effect of every parameter on the nature of the asymptotic
regime given the model used.
TIP5-CT-2006-031312 Page 34 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.3.2. Theoretical bases for the study
(a) WORKING HYPOTHESES
We studied a rail subject to repeated loading by a train wheel. The component material of the rail is
considered to be homogenous and isothermal at all times. It is modelled by an elastoplastic material. We
worked in the context of small disruptions (HPP). Several material behaviour laws may be considered;
we will use Von Mises type criteria, which are well adapted to metals.
We worked in 2 dimensions first of all. The rail is seen as a semi-infinite solid object. We will return to
this hypothesis during the parametric study
The wheel/rail contact force is modelled using Hertzs theory. The normal force is given by
2
max 2
( ) 1
x
p x p
b
= where pmax is the normal maximum pressure and b the half-length of the contact.
The tangential force is equal to f * p(x).
Vertical movement is blocked on the bottom of the rail.
Here is the list of parameters that in principle enable us to define our problem:
Young modulus E
Poissons ratio v
Density
Elasticity limit in simple traction oY
Hardening law and associated parameters (hardening modulus, etc.)
Half-length of the wheel-rail contact b
Normal maximum pressure pmax
Coefficient of friction f
Train speed V
Since all the rails are made of steel, we choose to set E at 210GPa and v at 0.3.
(b) THE STATIONARY METHOD
To model the passage of a wheel over the rail in the standard manner, we must resolve a dynamic
problem by having the elliptic force moved from one end of the rail to the other, using time steps. We
must repeat this operation until we find the asymptotic response of the rail. This would require a
considerable calculation time.
TIP5-CT-2006-031312 Page 35 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Figure 5.1.18: Stationary problem
For structures subject to a moving load at a fixed speed, a method called the stationary method was
developed by MM. Dang Van and Maitournam [1 to 4]. It applies perfectly to the case being studied of an
elliptic load on a rail. It involves going to the marker related to the load and adding a volume drive
inertia force to the material. The advantage lies in removing the dependency on time in the equations. In
fact, the temporal drift of a vector A is written
.
A A
V gradA V
t x
c c
= =
c c
The problem is then resolved by imposing a static force on the rail and by resolving the problems
equations from the right to the left. The initial conditions in terms of plastic deformation and internal
variables are given by the plastic deformations and internal variables in x = +.
The problems equations are the following:
Balance
2
, xx
div V u o =
in O
Boundary conditions .
d
n T o = on cST
d
u u =
on cSu
Initial conditions ( ) ( )
0
,
p p
y y c c + =
( ) ( )
0
,
k k
y y + =
Behaviour
el p
c c c = +
0
:
el
L o o c = +
,
p
x
f o
c
oo
=
,

A
k x
k
f o

o
=
f, ok and Ak depend on the hardening model selected. Nevertheless, we can provide a general form for
the plasticity criterion f:
( )
, ( )
k Y
eq
f X R p o o o = where X is the centre of the elasticity zone,
depending on ok, and p the cumulated plastic deformation.
(c) ADDITIONAL HYPOTHESES
It is possible to carry out a few additional approximations that will simplify our study. To do so, let us
examine the variation formulation of the elastic problem:
Find u e C(u
d
) so that for any field w e C(0) we have
0, f(o, o
k
)0, and f = 0
TIP5-CT-2006-031312 Page 36 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
2
2
2
( ) : : ( ) . .
T
d
S
u
u L w dV T wdS V wdV
x
c c
O O
c
= +
c
} } }
and compare the scales of the terms in the first and last integral thanks to the ratio r (Cs denotes the
velocity of the mechanical waves in steel, h the characteristic size of the mesh and Ax the typical distance
between 2 columns, see 5.1.3.3(a)).
2 2
2
2 2 2
2
2
( ) : : ( )
.
.
s
u L w
C E h x
r
u V x V h
V w
x
c c

A | | | |
=
| |
c A
\ . \ .
c

But Ax~ h, so ( )
2
s
r C V and as Cs ~ 6000km/h for steel, we have r >>1.
We therefore choose to deal with the problem quasi-statically. The variation formulation becomes
Find u e C(u
d
) so that for any field w e C(0) we have
( ) : : ( ) .
T
d
S
u L w dV T wdS c c
O
=
} }
We find the usual formulation. The difference is that the problem is resolved numerically the closer from
the right to the left in the mesh (see 5.1.II.1).
In addition we suppose that the plastic deformations and the initial constraints in the rail are zero.
However, the algorithms written under Castem provide the user with possibility of entering fields that
are not zero.
We use a cinematic linear hardening law. The plasticity criterion is
Y
f X o o = with
p
X Hc = .
The hardening modulus H is a parameter in our study.
We therefore give the parameters retained for our study and the interval of the values on which we work
and which correspond to realistic values. We will see in 5.1.II.1. that we must add a parameter related to
the mesh.
Parameter H (GPa) SigmaY tensile
strength(MPa)
Pmax (Mpa) f b (mm)
minimum 20 200 300 -0.6 2
maximum 100 900 2000 0.6 6
Table 5.1.10: Range of parameters
(d) CALCULATIONMETHODS
The stationary method requires specific meshing that limits the rail in vertical columns numbered from 1
to N from right to left.
The algorithm carries out several elastoplastic calculations in order to find the stationary solution. A
calculation is performed as follows:
The plastic deformations and the internal variables are given in column 1 (x = +) either by the user in
the first calculation or via the result of the previous calculations.
TIP5-CT-2006-031312 Page 37 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
For column n+1>1, the plasticity criterion is calculated at each Gauss point. If it is respected (f < 0) we
have
, 1 ,
1
,
p p
k n k n
n n
c c o o
+
+
= = . If the criterion is not met (f > 0), we calculate the increment in the plastic
deformations and the internal variables with the behaviour laws in I.2.; we have
, 1 , ,
1
,
p p p
k n k n k n
n n n
c c c o o o
+
+
= + A = + A . See [1] [2] and [11] for more details.
The algorithm converges when 2 stoppage conditions are verified:
The plasticity criterion is respected at all points.
The plastic deformations are stationary. This means that we find the same plastic status upstream and
downstream of the wheel: ( ) ( )
p p
x x c c = = = +
We must then specify what the algorithm does between 2 elastoplastic calculations. There are two
possibilities.
(i) Passage by passage method (indirect)
This method calculates the result of each wheel passage.
Each passage may require several elastoplastic calculations between which we conserve the previous
plastic deformation field then carry out a new elastic prediction. The plastic deformations in the column 1
in calculation i+1 are those found in calculation i in this same column. The operation is repeated until the
plasticity criterion is respected everywhere.
Then, between two passages, we recover the plastic deformation in x = - and impose it in all the other
meshing columns. We therefore have the rail as it actually is between 2 passages. Then we carry out a
new elastic prediction, which is the starting point for the calculation of the next passage. The algorithm
converges when the plastic deformations in x = - are identical between 2 successive wheel passages.
This method has the advantage that it is exact outside the approximation of the finite elements. The proof
was provided by MM. Dang Van and Maitournam [1]. However, it does require a certain calculation
time.
(ii) Direct method
This method does not differentiate between the wheel passages.
It involves carrying out the elastoplastic calculations in succession, but this time by imposing in x = +
for the calculation i+1 the plastic deformations found in x = - at the end of calculation i.
At the end of each calculation we check to see whether the plastic deformations x = - and in x = + are
identical. If this is the case we then check the plasticity criterion. If this criterion is respected everywhere,
the algorithm converges. If either of the two stoppage conditions is not checked, then we move onto the
next calculation.
This method requires reduced calculation time. It leads to the same result as the indirect method
concerning the nature of the asymptotic regime. However, the scale of the plastic deformations and other
solution fields is not the same as with the indirect method whose accuracy has been demonstrated.
TIP5-CT-2006-031312 Page 38 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(iii) Asymptotic regime
Once the stationary solution has been found with one of the methods above, we only need to determine
the nature of the asymptotic regime achieved.
The case of a purely elastic regime is clear since the plasticity criterion is checked from the start. The
algorithm does not carry out an elastoplastic calculation.
For elastic shakedown and plastic shakedown, we take the plastic deformations' solution field, extract the
plastic deformations from column 1 and then compare them with the plastic deformations in the other
columns. If they are still identical, this means that they do not change during a load cycle; this is elastic
shakedown. If there is at least one column for which they vary, this means that they change during a load
cycle but finish in the same state as at the start of the cycle; this is plastic shakedown.
TIP5-CT-2006-031312 Page 39 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.3.3. Implementation under Castem
(a) GENERAL
Castem, also called Cast3M, is a partial equation resolution application that uses finite elements
developed by the CEA.
Castem is able to mesh a structure whose geometry is defined by the user. It also carries out post-
processing for results.
It is a very open code which is well suited to research work. Nevertheless, the interface with the user is
almost zero since it does not have its own editor, it is executed in a command window and is not simple
to use.
(b) MESHING
To represent a portion of rail in 2D, we use a simple rectangle. We must then choose several things: the
dimensions of this rectangle, the size of the elements to ensure that the meshing is not too rough and the
type of element.
Since we must create a meshing composed of vertical columns, we opt for 8-point rectangular elements
(QUA8). We could have taken 4-point elements only (QUA4).
The piece of rail modelled must be large enough in relation to the size of the wheel-rail contact (b). At the
same time, the meshing must be fine enough under the contact zone.
Initially, we used the following meshing (the load is also represented, the different sizes of the nodal
forces are related to the use of QUA8 elements):
Figure 5.1.19: Initial meshing and load
b
5b
2b
3b
x
y
TIP5-CT-2006-031312 Page 40 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
This meshing was sufficiently fine but not large enough, which altered the results. In addition, we need
independent meshing for parameter b to be able study the effect of b on the nature of the asymptotic
regime. This leads us to introduce a new parameter that we note a. To avoid overloading the computers
memory, we have created a meshing with variable density, which is larger but which keeps almost the
same number of elements.
Figure 5.1.20: Meshing used and load
This meshing contains 100 columns with 50 elements each.
We can already see that b and a must be relatively close. We will return to this in the parametric study.
(i) Boundary conditions
As indicated in 5.1.I.1., the movement is blocked according to y for the P2P3 straight line. To enable the
computer to find a solution (uniqueness problem) the structure must be blocked according to x. Several
solutions may be considered: block the straight line P2P3 according to x (and still y), block the straight
line P1P2 according to x, or block point P2 according to x. These alternatives are discussed in (c)(iii).
(c) SPECIFIC PROCEDURES
(i) Plastic increment
Lets suppose that we are at a point in the calculation where one of the Gauss points does not respect the
plasticity criterion. We must then calculate the increment in the plastic deformations and internal
variables for this point by a radial return algorithm.
There are two problems with Castem. It is difficult to access the Gauss points individually; we must
handle the fields defined over a column and there is no radial return function included. Two similar
procedures are provided in Castem but they are not very well documented and it is difficult to find out
what they calculate effectively. This is why we have preferred to write a radial return algorithm
ourselves. We may describe it as follows:
2b
3a
12a
4a
13a
P1
P3
P4
P2
x
y
TIP5-CT-2006-031312 Page 41 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
The entry fields are the characteristics of the material, the elastic prediction for the deformations c
las
, the
plastic deformations and internal variables in the previous column. For the cinematic linear hardening,
the computer calculates:
las prd p
c c c =
( )
0
( ) 2
las las las
S dev tr o c c = + + 1 , then
1
2
2
3
Y
p
las
S
las
S
las H
S
c o

| |
| A =
+
\ .
Which enables us to update the plastic deformations and internal variables.
The simultaneous handling for all the points in the column makes the coding of these relatively simple
calculations quite delicate.
(ii) Stoppage tests
Let us return to the operation of the stoppage tests.
c.ii-1 Steady state
This involves comparing two deformation fields c
p
1 and c
p
2 defined on the same geometric support (one
column) with a certain tolerance tol, typically 1%. For this, we start by setting to 0 the deformations lower
than 10
-5
. This enables us to avoid any problems at points with a relatively low level of deformation. Then
we calculate
1 2
p p
c c which we will compare to a reference value. This is defined by
( )
1
2
1 2
max max
p p
tol c c + . If there is a point where the difference in the two fields is greater than
the reference value, the test is false.
c.ii-2 Plasticity criterion
On input we provide the characteristics of the material, an internal variable field (X) and a tolerance tol.
We create a reference value equal to '
Y
tol o , we calculate the criterion then we check at each point that
'
Y Y
f X tol o o o = s . If a point violates this equality, the test is false.
(iii) Limits
c.iii-1 Boundary conditions
The Boundary conditions in movement have a significant impact on the program's behaviour. Here we
discuss the different possibilities for the blockage according to x.
TIP5-CT-2006-031312 Page 42 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Figure 5.1.21: Meshing and its deformation; P2P3 blocked according to UX and UY.
Figure 5.1.222: Meshing and its deformation; P2P3 blocked according to UY, P2 blocked according to UX.
Figure 5.1.33: Meshing and its deformation; P2P3 blocked according to UY, P1P2 blocked according to UX.
If we block the movement according to y and x for the straight line P2P3 we quickly encounter edging
problems. In fact, points P2 and P3 are places where stresses are concentrated and they plastify when the
load is increased, which leads to undesirable behaviour for the program bad asymptotic regime at
medium load, non convergence at high load.
Blocking P2 according to x is a good alternative. We can see in figure 7 that the structure is deformed
significantly more according to x than on the previous figure. However, we encounter problems when we
TIP5-CT-2006-031312 Page 43 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
take non-zero tangential stresses. In fact, it is only point P2, which "retains" the structure. The stresses are
high at this point and for f > 0.2 it plasticizes well before the part under the load plasticizes, as shown
below.
Figure 5.1.24: Effect of the conditions on the limits (plastic deformations)
To get around this problem, we choose to block the whole straight line P1P2 according to x. The
deformations in figures 7 and 8 are very close, but the stresses are highly distributed and are no longer
concentrated on P2. Except otherwise stated, this boundary condition has been used.
In addition, in order to avoid other edge effects, we exclude the two columns for each end. This means
that if there are N columns we assimilate x = + in column 3 and x = - in column (N 2).
c.iii-2 Accuracy
The two stoppage tests have a certain tolerance set by the user. Initially the tolerances were equal. This
does not cause any problems with the direct method since exactly the same calculation is repeated until
the two tests are verified. However, the two tolerances each play a different role in the indirect method.
The tolerance tol of the plasticity criterion determines how accurate a passage is. With low tolerance, the
calculation of a passage requires more time but is more accurate and, all things being equal, the algorithm
needs to simulate fewer passages to converge.
The tolerance tol of the stationary aspect tests enables the calculation to be stopped when the plastic
deformations from one passage to another are close enough. With small tolerance, the algorithm needs to
simulate more passages.
Finally, we must choose a small tol and an tol that is not too small to avoid a long calculation time, but
still small enough to be sure we have reached a stationary state. We estimate that tol = 1% is an acceptable
compromise. From this point, we have tested several values for tol. In the end we have retained
tol=0.5%. We illustrate this by tracing the maximum of the difference in plastic deformations in -
between successive passages, using several tolerance pairs (in the legend, the first tolerance is the one for
the plasticity criterion, the second is the one for the stationary status test).
H (GPa) SigmaY tensile strength(MPa) Pmax (Mpa) f a (mm) b (mm)
70 500 1170 0 3 4
TIP5-CT-2006-031312 Page 44 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Table 5.1.11: Parameters
Figure 5.1.25: Maximum in inter-passage plastic deformation difference according to the passage for several
tolerance pairs.
We see that with the pair (tol=0.5%; tol = 1%) we converge in few passages and very quickly approach a
stationary solutions (the blue curve is below the green one).
TIP5-CT-2006-031312 Page 45 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.3.4. Study results
(a) ELASTIC SHAKEDOWN EXAMPLE
We present the plastic deformation results for an elastic shakedown case. Remember that we use the
meshing in figure 5. The passage-by-passage method was used.
H (GPa) SigmaY tensile strength(MPa) Pmax (Mpa) f a (mm) b (mm)
70 500 1000 0 3 4
Table 5.1.12: Parameters
Figure 5.1.26: XX plastic deformations
Figure 5.1.27: YY plastic deformations
Figure 5.1.28: XY plastic deformations
Figure 5.1.29:Plastic deformation standard
The plastic deformations do not change over a cycle; they are constant along the horizontal line.
TIP5-CT-2006-031312 Page 46 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(b) PLASTIC SHAKEDOWN EXAMPLE
The plastic shakedown case is richer in terms of interpretation. We only represent certain components of
the fields that we feel are the most relevant. The graphs for all the components for this elastic shakedown
case are provided in the appendix. They were obtained by the passage-by-passage method.
H (GPa) SigmaY tensile strength(MPa) Pmax (Mpa) f a (mm) b (mm)
70 500 1250 0 3 4
Table 5.1.13: Parameters
Figure 5.1.30: Equivalent plastic deformation
Figure 5.1.31: XY plastic deformations
The algorithm converges after 8 passages. We can clearly see under the wheel-rail contact area that the
plastic deformations change.
To better see how the algorithm works, we trace the plastic deformations for the passages 1, 2 and 8. We
trace them in x = - according to the depth and also at fixed depth (0.7b approx) according to x.
Figure 5.1.32: YY plastic deformations in x = -
Figure 5.1.33: XY plastic deformations in x = -
TIP5-CT-2006-031312 Page 47 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Figure 5.1.4: XX plastic deformations at fixed depth
Figure 5.1.35: YY plastic deformations at fixed depth
Figure 5.1.5: XY plastic deformations at fixed depth
Figure 5.1.37: Standard for plastic deformations at fixed
depth
We see that on the first passage the plastic deformations are zero upstream of the wheel. Passage 2 starts
in x = + with the plastic deformations found on passage 1 in x= - . For the stationary response, we find
the same plastic deformations upstream and downstream of the wheel. The algorithm behaves as
planned.
We note that the plastic deformations XY change symbol according to depth.
Finally, we look at the residual stresses in the rail after passage of the train in stationary state. To do so,
we trace the residual stresses in x = - according to the depth. They are at a maximum around y = -b.
TIP5-CT-2006-031312 Page 48 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Figure 5.1.38: XX residual constraints in x= -
Figure 5.1.39: YY residual constraints in x= -
Figure 5.1.6: XY residual constraints in x=
Figure 5.1.7: Equivalent residual constraint in x=
We may also trace the stress- strain curve to observe the load cycles. We see that the first cycles are open,
and then converge on a closed cycle, showing a stationary state
Figure 5.1.42: XY constraints XY deformations curve
Figure 5.1.43: XY constraints xy plastic deformations
curve
TIP5-CT-2006-031312 Page 49 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(c) PARAMETRIC STUDY
This involves determining the influence of different parameters on the nature of the asymptotic regime.
Let us start by recalling the parameters selected for this study.
Material parameters
o Elasticity limit in simple traction oY
o Hardening modulus H
Load parameters
o Half-length of the wheel-rail contact b
o Normal maximum pressure Pmax
o Coefficient of friction f
Parameter related to modelling
o Meshing parameter a
We give these parameters default values, which are used in the calculations, except where stated to the
contrary. These are the values for the elastic shakedown example 5.1.(a).
H (GPa) SigmaY tensile strength(MPa) Pmax (Mpa) f a (mm) b (mm)
70 500 1000 0 3 4
Table 5.1.14: Parameters fo elastic shakedown
TIP5-CT-2006-031312 Page 50 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(i) Load factor
We define the load factor as the ratio Pmax/oY. Let us show the importance of this factor by tracing the
elasticity and elastic shakedown limits in a diagram Pmax - oY.
Figure 5.1.44: Diagram Pmax - sY
We obtain straight lines, which shows the usefulness of the Pmax/oY ration to define the limits of the
different asymptotic behaviours.
This might let us think that there are two values for the load factor, which define the elasticity and elastic
shakedown limits (here 1.7 and 2.3). However, the other parameters may modify these threshold values.
Note that the two calculation methods provide the same limits. This result was never proved wrong
throughout the study.
Plastic shakedown or
Ratchetting
Elastic shakedown
Elasticity
TIP5-CT-2006-031312 Page 51 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii) Hardening modulus
Let us study the influence of the hardening modulus H. To do so we will set SigmaY at 500Mpa and vary
H.
Figure 5.1.45: Load factor hardening module diagram
The hardening modulus has no notable influence on the elasticity and elastic shakedown limits.
Elasticity
Elastic shakedown
Plastic shakedown or Ratchetting
TIP5-CT-2006-031312 Page 52 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(iii) Coefficient of friction
Here we study the effects of the tangential stresses.
Figure 5.1.468: Load factor rubbing coefficient diagram
We can see that the curve is not exactly symmetrical in relation to the y-axis. This can be explained simply
by the fact that our model is not completely symmetrical since we chose to block movement according to
x for the straight line P1-P2 located to the right of the mesh.
Concerning our study, the coefficient of friction lowers the elasticity and elastic shakedown limits. The
phenomenon is low for around |f|<0.25 environ. It is very high beyond this.
This means that for a rail located in a braking or acceleration area, we may easily obtain plastic
shakedown, which will accelerate fatigue on the rail.
Elasticity
Elastic shakedown
Plastic shakedown or Ratchetting
TIP5-CT-2006-031312 Page 53 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(iv) Size of the wheel-rail contact
First of all, we vary b only by keeping the mesh presented in 5.1.3.3(a).
Figure 5.1.47: Load factor size of the wheel-rail contact (usual meshing) diagram)
In principle, it is difficult to interpret these curves. In fact, reducing the contact surface seems to increase
the elasticity limit, which is counter-intuitive. The explanation is related to the mesh used: at the contact
level the mesh is denser, the meshes have a size of 0.3 mm over a length of 6 mm. But this is only adapted
if b is close to a. To show this, we have traced the same curves but by modifying each time to keep a=b.
Figure 5.1.48: Load factor size of the wheel-rail contact (a=b) diagram
TIP5-CT-2006-031312 Page 54 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
The dominant effect is the fineness of the meshing. To study the influence of the contact surface properly,
there must be meshing for which the densest part has meshes of characteristic size bmin/10 over a
characteristic distance 2bmax . This is what we have done, taking bmin=2mm and bmax=6mm.
Figure 5.1.49: Adapted meshing
This mesh contains 140 columns of 100 elements each. The calculations are much longer and we come
close to the limits of the computers memory.
Here are the results obtained.
Figure 5.1.50: Load factor size of the wheel-rail contact (adapted meshing) diagram
The size of the contact does not have a major influence on the nature of the asymptotic regime. This is
normal as we model a rail that has an infinite length. Regardless of the size of the contact, there is always
Hertz pressure applied to a semi-infinite solid object.
TIP5-CT-2006-031312 Page 55 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Things may be different in 3D since the ratio between the width of the rail and the width of the contact do
not necessarily enable us to apply the semi-infinite solid object hypothesis.
If we imagine the contact on the side depths of the rail, the results will be modified.
Nevertheless, the size of the contact is the characteristic size of the problem. For example, it modifies the
depth of the plastic deformations
Figure 5.1.51: Depth of the maximum plastic deformation according to the size of the wheel-rail contact
This curve has been traced with a load factor of 2.2 (elastic shakedown).
The maximum plastic deformation is found at a depth of around 0.64 b. The depth interval on which the
rail plasticizes also depends on the size of the contact. It does not exceed a depth of 3b.
TIP5-CT-2006-031312 Page 56 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(d) LIMITS
We have used a simple hardening model that has the advantage that is can be programmed easily and
only has a single parameter (H) but does have disadvantages.
For a cinematic, linear hardening, we cannot simulate the plastic ratchetting phenomenon (Halphen,
1977); the algorithm always converges at worst on a plastic shakedown case.
This type of model is only valid if the load is not too high. For a load factor of over 2.5, we obtain results
that are not very convincing, which we illustrate with the plastic deformations graph for a load factor of
3.4.
Figure 5.1.52: YY plastic deformations for a loaf factor of 3.4
Concerning the behaviour of the program with the model, the calculation time and the number of
passages for the indirect method on convergence do not follow a regular change. The algorithm seems to
converge in particular with more difficulty just after the plastic shakedown limit.
Material E (GPa) Nu H (GPa) SigmaY tensile strength
(MPa)
f a (mm) b (mm)
1 210 0.3 70 500 0 3 4
2 210 0.3 90 300 0 3 4
Table 5.1.15: Parameters value
Figure 5.1.53: Calculation time according to the load
factor
Figure 5.1.54: Number of passages simulated by the
indirect method according to the load
factor
TIP5-CT-2006-031312 Page 57 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(e) FUTURE DEVELOPMENTS
First of all, we want to set up a more realistic material model. We have done this for a cinematic linear
and isotropic non-linear hardening combined, but the algorithm no longer converges as soon as we reach
the elastic hardening limit. Nevertheless, this limit is obtained when the load factor is 2.3. This lets us
think that the hardening law has no influence on the elastic shakedown limit.
Then we will move to a 3D model, which will enable us to modify the form of the wheel-rail contact
(circular, elliptic, etc.) and to consider a contact on the depth of the rail. This will be done during July.
The longer-term objective is to use the stationary method with the ANSYS application and to study
fatigue on the rail by using the Dang Van criterion in particular.
TIP5-CT-2006-031312 Page 58 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.3.5. Conclusion
We have presented the stationary method and its implementation. Under certain hypotheses, this method
has enabled us to carry out a parametric study. Let us summarise the main results.
For a cinematic linear hardening, in 2D, in the quasi-static context, the nature of the asymptotic regime is
determined by:
The load factor, equal to the ratio Pmax/oY
The rubbing coefficient
For a zero rubbing coefficient, the asymptotic regime is the following:
Elasticity for a load factor lower than 1.7
Elastic shakedown for a load factor between 1.7 and 2.3
Plastic shakedown for a load factor over 2.3
The hardening modulus and the size of the wheel-rail contact has no significant influence.
The mesh must be adapted to the size of the wheel-rail contact.
The rail fatigue study has two practical advantages: to be able to make a commitment on the lifetime of
the rails installed and to optimise the rail replacement frequencies for the track maintenance.
TIP5-CT-2006-031312 Page 59 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.3.6. Bibliographical references
(i) Stationary method and rail fatigue
[1] K. Dang Van, M.H. Maitournam; Steady-state flow in classical elastoplasticity: applications to
repeated rolling and sliding contact, J.Mech.Phys.Solids Vol.41 No11, 1691-1710 (1993).
[2] M.H. Maitournam, Formulation et resolution numrique des problmes thermoviscoplastiquesen
regime permanent, PhD thesis, Ecole Nationale des Ponts et Chausses, Paris, France (1989).
[3] K. Dang Van, M.H. Maitournam, Elastoplastic analysis of repeated moving contact: Application to
railways damage phenomena, Wear 196, 77-81 (1996).
[4] K. Dang Van, M.H. Maitournam, On some recent trends in modelling of contact fatigue and wear in
rail, Wear 253, 219-227 (2002).
[5] K. Dang Van, Modelling of damage induced by contacts between solids, C.R. Mcanique 336, 91-101
(2008).
[6] K. Dang Van, M.H. Maitournam, Z. Moumni, F. Roger, A global approach for modeling fatigue and
fracture of rails, Cinquimes Journes dEtudes Techniques 2008, Marrakech, Maroc (2008).
[7] N. Maouche, M.H. Maitournam, K. Dang Van, On a new method of evaluation of the inelastic state
due to moving contacts, Wear 203-204 , 139-147 (1997).
(ii) Material mechanics
[9] M.H. Maitournam, Mcanique des structures anlastiques, Ecole Polytechnique, ISBN 978-2-7302-
1543-5 (2009).
[10] P. Suquet, Rupture et plasticit, Ecole Polytechnique, ISBN 978-2-7302-1520-6 (2008).
(iii) Finite elements and Castem documentation
[11] M Bonnet, A. Frangi, Analyse des solides dformables par la mthode des lments finis, Les
Editions de lEcole Polytechnique, ISBN 978-2-7302-1349-3 (2007).
[12] http://www-cast3m.cea.fr/cast3m/index.jsp notamment Utilisateurs\Notices et
Utilisateurs\Documentation.
[13] http://rdforum.free.fr/cast3m/ notably Installation, configuration, launch under Windows.
TIP5-CT-2006-031312 Page 60 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.4 Measurements in Madrid
5.1.4.1. Introduction
Noise and vibration measurements have been performed between 19 and 20 May 2010 on the Metro
Ligero Oeste line Boadilla, in order to compare three CDM systems, which are.
Classic + Manta (S1): CDM-QTrack-HP + FST on mat FSM-L13;
Classic (S2): CDM-QTrack-HP;
Comfort (S3): QTrack-XP.
The following measurements have been performed for all systems:
Rail admittance measurements;
Rail roughness measurements;
Pass-by measurements: noise and vibrations.
The Technical Report attached in Appendix provides all the details of this measurement campaign.
5.1.4.2. Test Conditions
(a) TEST VEHICLE
The test vehicle is the ALSTOM CITADIS 302.
Figure 5.1.55: Citadis 302 in Madrid
TIP5-CT-2006-031312 Page 61 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(i) Metro Ligero Oeste vehicles
The Metro Ligero Oeste vehicles, model CITADIS 302, were manufactured by ALSTOM. They can carry nearly
200 passengers. They run on clean electricity, they are extremely quiet and have an extremely luminous interior
design.
Metro Ligero Oeste S.A. now has 27 of the 70 CITADIS 302 vehicles, bought from ALSTOM in the first stage by
MINTRA.
These are amongst the most modern vehicles in the world at present. They run in tracks on electric-driven steel
wheels, they have an excellent braking system; they are easy to board (100% low floor), with four double doors and
wide aisles to hold wheelchairs and bicycles.
Although light rail vehicles mostly run on their own rights-of-way, in the case of Metro Ligero Oeste, they are also
designed to share their space with buses, cars, cyclists and pedestrians. They provide frequent services and have
priority over the rest of the traffic.
(ii) Technical details of light rail vehicles
Total length 32.34m
Total width 2.40 m
Total height 3.60 m
Entrance height 320 mm
Track gauge 1.439m
Composite vehicle weight 39.9 tons
Number of seated passengers 54
Standing capacity 132
Total capacity 186
Power 4 x 120kW
Power Supply: 750 V cc / 400 V ac / 24 V dc
(b) TRACK SYSTEMTYPES
In three different sections, three different systems are measured:
Classic + Manta (S1): CDM-QTrack-HP + FST on mat FSM-L13;
Classic (S2): CDM-QTrack-HP;
Comfort (S3): QTrack-XP.
The measurement sites are situated on the line from Puerta de Boadilla to Colonia Jardin. The location of
the measurement sites is illustrated in the following figures, showing two maps and one photograph per
section/system.
TIP5-CT-2006-031312 Page 62 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Figure 5.1.56: Map of system Classic + Manta
Figure 5.1.57: Map of system Classic
Figure 5.1.58: Map of system Comfort
TIP5-CT-2006-031312 Page 63 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.4.3. Geometry Measurement
(a) INTERPRETATION OF READINGS
(i) Measurement processing
The track irregularities measurement trolley (called Krab-Light trolley and illustrated hereunder) allows
plotting 5 preliminary signals, which constitute the basis for post-processing. These 5 preliminary signals
are:
The preliminary gauge with a 1435mm nominal gauge basis
The preliminary cant measured thanks to an inclinometer (positive at right, negative at left).
The preliminary vertical alignment measured on the right of the trolley on an 1.9m asymmetric basis
(positive at left, negative at right)
The preliminary horizontal alignment on the right side and on the same basis.
The preliminary twist with a twist-measuring pole.
Figure 5.1.59: Krab-Light trolley
(ii) Post-processed measurements
A post-processing is applied and allows bringing back the track irregularities to a workable basis.
Characteristics of this post-processing is defined by EN 13848 standards.
These European standards require to process these data using a ButterWorth type band-pass filter, in a
wavelength range between 2 and 25m.
During post-processing, the software uses the 5 preliminary signals and the wavelength parameter
defined by EN 13848 standards to obtain:
A quasi-static signal (low-frequency or high-wavelength signal).
A dynamic signal (high-frequency or low-wavelength signal).
An overall signal (sum of both signals).
The post-processing allows obtaining the following data for each track parameter:
Measurement of
preliminary vertical
alignment and horizontal
alignment
Measurement
of preliminary
twist
TIP5-CT-2006-031312 Page 64 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Table 5.1.16: Post-processed parameters
The post-processing calculates the following parameters for each filtered track irregularity:
The quasi-static signal is a mean line, obtained by a least squares method. This signal is close to the
designed signal.
The dynamic signal is obtained calculating the gap between the quasi-static signal and the
unprocessed signal.
The overall signal is the sum of the quasi-static signal and the dynamic signal.
The vertical alignment and the horizontal alignment are brought to a 10m asymmetric basis,
according to European standards.
The curvature is the track deflection measured on a 10m symmetric chord.
The twist is calculated on 3 different basis, in accordance with the operator choice, with a maximum
value of 20m (for Madrid measurement: 3m, 5m and 10m)
On the overall post-processed parameters, the following ones are studied:
The gauge
The dynamic signal of the cant
The dynamic signal of the vertical alignment measured at the centre of the track on a 10m basis
The horizontal alignment measured at the track centre axis on a 10m basis
The twist on a 3m basis.
The axis curvature (Curvature = 1000/R where R = arc radius)
Note: The design gauge for the ML-3 line is 1435 mm
These post-processed parameters are analysed as follows:
A statistic study per 200m sections
A study in curves.
TIP5-CT-2006-031312 Page 65 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(b) STATISTIC STUDY OF V1: [PUERTADEBOADILLA-COLONIA JARDIN]
(i) Principle:
The statistic study is based on the previous numeric results to generate an analysis by 200m sections.
Thus, for each section and for each geometric parameter, the mean value, the maximum value, the
minimum value and the standard deviation are calculated.
The direction of travel for measurements on track V1 is shown hereafter:
Figure 5.1.60: Direction of travel for measurements - V1
(ii) Gauge
The following figure presents the standard deviation for the gauge per 200m section:
Figure 5.1.61: Standard deviation for gauge per 200m section - V1
b.ii-1 General observations
We note that, on the overall length of the track V1, the mean standard deviation is 1.3mm. Thus, the track
gauge is 1435mm +/- 2.6mm for a 95% compliance and 1435mm +/- 3.9mm for a 99% compliance.
However, on all sections analysed and considering a 95% compliance, 12 no's of 200m sections are not
compliant with the EN 13231-1 standard. In fact, the gauge in these sections is higher than +/-3mm.
We will study these sections in order to define the reasons of this non-compliance.
Direction of travel for measurement V1
95% of
compliance
99% of
compliance
TIP5-CT-2006-031312 Page 66 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.ii-2 Non-compliance analysis
On non-compliant sections, we have selected 2 representative sections: the section [2,4: 2,6[ and the
section [9,6: 9,8[.
b.ii-2.1 Section [2,4: 2,6[:
Figure 5.1.62: Gauge defects in section [2,4: 2,6[
We note that the track gauge exceeds the EN 13231-1 standard acceptance tolerances in a small radius
curve (R=60m) as well as in a straight alignment. The first defect is related to a wear phenomenon in
small radius curves. The second one may be related to a track installation defect (for information, this
section is a descendant slope).
b.ii-2.2 Section [9,6: 9,8[:
Figure 5.1.63: Gauge defects in section [9,6: 9,8[
We note that the track gauge exceeds the acceptance tolerances in a small radius curve (R=40m). This
defect is related to a wear phenomenon in small radius curves.
TIP5-CT-2006-031312 Page 67 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(iii) Vertical alignement
The following figure presents the standard deviation for the vertical alignment per 200m section:
Figure 5.1.64: Standard deviation for vertical alignment per 200m section - V1
b.iii-1 General observations
We note that, on the overall length of the track V1, the mean standard deviation for vertical alignment is
1.3mm. Thus, the vertical alignment range is between 2.6mm and +2.6mm for a 95% compliance and
between 3.9mm and +3.9mm for a 99% compliance.
However, considering a 95% compliance, we note in a 200m section a standard deviation higher than
4mm. So, this section is non-compliant with the EN 13231-1 standard acceptance tolerances for vertical
alignment.
We will study this section in order to define the reasons of this non-compliance.
95% of
compliance
99% of
compliance
TIP5-CT-2006-031312 Page 68 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.iii-2 Non-compliance analysis
b.iii-2.1 Section [2,8: 3[:
Figure 5.1.65: Vertical alignment defects in section [2,4: 2,6[
We note that the vertical alignment defects exceeds the acceptance tolerances in a small radius curve
(R=50m). This defect is not related to a track irregularity but it is related to a post-processing fault.
In addition, we will study the section [9,6: 9,8[, which is compliant for the vertical alignment but not for
gauge, cant and twist
b.iii-2.2 Section [9,6: 9,8[:
Figure 5.1.66: Vertical alignment defects in section [9,6: 9,8[
We note that the vertical alignment oscillates a lot around its mean value. This is due to a cant defect in
this section.
TIP5-CT-2006-031312 Page 69 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(iv) Twist
The following figure presents the standard deviation for the twist per 200m section:
Figure 5.1.67: Standard deviation for twist per 200m section - V1
b.iv-1 General observations
We note that, on the overall length of the track V1, the mean standard deviation for twist is 1.8mm. Thus,
the twist range is between 3.6mm and +3.6mm for a 95% compliance and between 5.4mm and +5.4mm
for a 99% compliance.
However, considering a 95% compliance, we note in a few sections a standard deviation higher than the
EN 13231-1 standard acceptance tolerances.
We will study these sections in order to define the reasons of this non-compliance.
95% of
compliance
99% of
compliance
TIP5-CT-2006-031312 Page 70 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.iv-2 Non-compliance analysis
On all non-compliant sections, we have selected 2 representative sections of the twist defects: the section
[2,6: 2,8[ and the section [9,4: 9,6[. In addition, we will analyse the section [9,6: 9,8[, which combines
several defects (gauge, cant and twist).
b.iv-2.1 Section [2,6: 2,8[:
Figure 5.1.68: Twist defects in section [2,6: 2,8[
We note that the twist defects are due to the variable applied cant in curve.
b.iv-2.2 Section [9,4: 9,6[:
Figure 5.1.69: Twist defects in section [9,4: 9,6[
We note that the twist defects are related to cant transition areas. Thus, in order to analyse deeper this
kind of defects, it is necessary to know the design twist in these cant transition areas, which will allow us
to highlight the real twist defect.
TIP5-CT-2006-031312 Page 71 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.iv-2.3 Section [9,6: 9,8[:
Figure 5.1.70: Twist defects in section [9,6: 9,8[
We note that the twist defects are related to cant transition areas. Moreover, these defects are also related
to a cant defect.
TIP5-CT-2006-031312 Page 72 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(v) Cant
The following figure presents the standard deviation for the cant per 200m section:
Figure 5.1.71: Standard deviation for cant per 200m section - V1
b.v-1 General observations
We note that, on the overall length of th track V1, the mean standard deviation is 1mm. Thus, the cant
defect range is between 2mm et +2mm for a 95% compliance and between 3mm et +3mm for a 99%
compliance
However, considering a 95% compliance,we note that, in a 200m section, the standard deviation is
around 3mm. This value is higher than the acceptance tolerance of the EN 13231-1.
We will study this section in order to define the reasons of this non-compliance.
95% of
compliance
99% of
compliance
TIP5-CT-2006-031312 Page 73 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.v-2 Non-compliance analysis
b.v-2.1 Section [9,6: 9,8[:
Figure 5.1.9: Cant defects in section [9,6: 9,8[
We note that the quasi-static cant (or applied cant) is inconstant in the first curve. In addition, in this
curve, the cant defect oscillates a lot, switching from a negative value to a positive value quickly. This can
represent a risk for safety.
TIP5-CT-2006-031312 Page 74 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(vi) Horizontal alignment
The following figure presents the standard deviation for the horizontal alignment per 200m section:
Figure 5.1.73: Standard deviation for horizontal alignment per 200m section - V1
b.vi-1 General observations
We note that, on the overall length of the track V1, the mean standard deviation is 1.8mm. Thus, the
horizontal alignment defect range is between 3.6mm et +3.6mm for a 95% compliance and between
5.4mm et +5.4mm for a 99% compliance.
However, considering a 95% compliance, we note that, in several sections, the standard deviation is
higher than the acceptance tolernace of the EN 13231-1 standard.
We will study this section in order to define the reasons of this non-compliance.
95% of
compliance
99% of
compliance
TIP5-CT-2006-031312 Page 75 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.vi-1.1 Non-compliance analysis
On all non-compliant sections, we have selected a representative section for this defect: the section [0:
0,2[.
Figure 5.1.74: Horizontal alignment defects in section [0: 0,2[
We note that the horizontal alignment exceeds the tolerances of the EN 13231-1 standard in small radius
curves. This defect is not due to a track irregularity but to a post-processing defect.
TIP5-CT-2006-031312 Page 76 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(vii) Compliance of the track V1 to the EN 13231-1 standard
For the overall length of the track V1 and for each parameter, the percentage of non-compliant track is
listed in the table below:
Geometric parameter Acceptance tolerances % of non-compliance
Gauge 3mm 4,93
Vertical alignment 6mm 0,39
Cant 3mm 1,64
Twist 4.5mm 5,17
Horizontal alignment 5mm 3,43
Table 5.1.17: Percentage of non-compliant track - V1
In order to highlight where these non-compliance are located, we will analyse these parameters in
relation to the curve radius. The table presented hereafter shows the results:
25 < R s 150 150 < R s 500 500 < R s 2000 R >2000 and AD
Geometric parameter Tolerances
% of non-
compliance
% of non-
compliance
% of non-
compliance
% of non-
compliance
Gauge 3mm 6,58 3,54 4,49 3,27
Vertical alignment 6mm 1,10 0,23 0,17 0,09
Cant 3mm 3,36 1,34 1,15 0,81
Twist 4.5mm 10,51 10,43 1,33 0,94
Horizontal alignment 5mm 12,46 0,42 0,30 0,01
Table 5.1.18: Percentage of non-compliant track per parameter and curve radius - V1
The gauge is the most sensitive parameter for concrete slab tracks (directly related to the wear
phenomenon in small radius curves), we will compare the repartition graph in relation to the curve
radius.
Figure 5.1.75: Comparison of gauge repartition graph in relation to the curve radius- V1
TIP5-CT-2006-031312 Page 77 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(viii) Conclusion
The percentage of non-compliant track for gauge, vertical alignment, and cant and horizontal alignment
is lower than 5%. Thus, considering 95% compliance, we conclude that the track V1 of the ML-3 line is
compliant to the EN 13231-1 standard acceptance tolerances for these track parameters. For the twist, the
percentage of non-compliant track is a bit higher than 5% but this value may not be the real twist defect.
In fact, we will be able to conclude for the compliance of this parameter when we will know the design
twist in cant transition areas (areas where we note many twist defects).
In addition, for the gauge, a part of the percentage of non-compliance can be related to the wear
phenomenon in small radius curves (for information, the ML-3 line of the Madrid tramway network was
opened to traffic in 2007). The rest of this percentage may be related to track installation issues and must
be highlighted. For the other parameters, the percentage of non-compliance is either due to a track
installation issue or due to a post-processing defect. So, in order to refine the compliance analysis, the
track design for each parameter must be an input data for post-processing.
According to the table 4, we can note that, in general, the percentage of non-compliant track increases
when the radius curve is decreasing. In addition, we note that for the horizontal alignment, the
percentage of non-compliant track is focused on the curve radius range between 25 and 150m and for the
twist; it is focused on the curve radius range between 25m and 500m.
According to the comparison of the gauge repartition graphs in relation to the curve radius, we can note
that, for straight alignment, the mean gauge is lower to the designed gauge (around 1434.5mm). We can
also note that this mean gauge increases when the curve radius is decreasing up to 1436mm for the curve
radius range between 25m and 150m.
Note that a particular attention must be paid to the section [9.6: 9.8[, near Retamares station, which
combine high track geometry defects for several parameters, and so, can represent of safety risk.
TIP5-CT-2006-031312 Page 78 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(c) STATISTIC STUDY OF V2: [PUERTA DE BOADILLA-COLONIA JARDIN]
(i) Principle:
The statistic study is based on the same principle than the statistic study of track V1. The non-compliance
analysis highlighting the same reasons than these of track V1 analysis, we will only remind these reasons
in this study.
The direction of travel for measurements on track V2 is shown hereafter:
Figure 5.1.76: Direction of travel for measurements - V2
(ii) Gauge
The following figure presents the standard deviation for the gauge per section of 200m:
Figure 5.1.77: Standard deviation for gauge per section of 200m V2
c.ii-1 General observations
We note that, on the overall length of the track V1, the mean standard deviation is 1.3mm. Thus, the track
gauge is 1435mm +/- 2.6mm for a 95% compliance and 1435mm +/- 3.9mm for a 99% compliance.
However, on all sections analysed and considering a 95% compliance, several sections are not compliant
with the EN 13231-1 standard. In fact, the gauge in these sections is higher than +/-3mm.
Direction of travel for measurement V2
95% of compliance
99% of compliance
TIP5-CT-2006-031312 Page 79 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Remider of the reasons of gauge defects for track V1
Defect related to a wear phenomenon in small radius curves.
Defect may be related to a track installation defect
(iii) Vertical alignement
The following figure presents the standard deviation for the vertical alignment per 200m section:
Figure 5.1.78: Standard deviation for vertical alignment per 200m section - V2
c.iii-1 General observations
We note that, on the overall length of the track V1, the mean standard deviation for vertical alignment is
1.3mm. Thus, the vertical alignment range is between 2.6mm and +2.6mm for a 95% compliance and
between 3.9mm and +3.9mm for a 99% compliance.
However, considering a 95% compliance, we note that, in a 200m section, the standard deviation is
around 3.5mm. So, this section is non-compliant with the EN 13231-1 standard acceptance tolerances for
vertical alignment.
Remider of the reasons of vertical alignment defects for track V1
Defect related to post-processing fault.
Defect due to a cant defect (non applicable in track V2)
95% of compliance
99% of compliance
TIP5-CT-2006-031312 Page 80 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(iv) Twist
The following figure presents the standard deviation for the twist per 200m section:
Figure 5.1.79: Standard deviation for twist per 200m section V2
c.iv-1 General observations
We note that, on the overall length of the track V1, the mean standard deviation for twist is 1.7mm. Thus,
the twist range is between 3.4mm and +3.4mm for a 95% compliance and between 5.1mm and +5.1mm
for a 99% compliance.
However, considering a 95% compliance, we note in a few sections a standard deviation higher than the
EN 13231-1 standard acceptance tolerances.
Remider of the reasons of twist defects for track V1
Defect due to a variable applied cant in curve.
Defect related to cant transition areas.
95% of compliance
99% of compliance
TIP5-CT-2006-031312 Page 81 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(v) Cant
The following figure presents the standard deviation for the cant per section of 200m:
Figure 5.1.80: Standard deviation for cant per 200m section V2
c.v-1 General observations
We note that, on the overall length of th track V1, the mean standard deviation is 1mm. Thus, the cant
defect range is between 2mm et +2mm for a 95% compliance and between 3mm et +3mm for a 99%
compliance
However, considering a 95% compliance,we note that, in 2 no's of 200m section, the standard deviation
exceeds the acceptance tolerance of the EN 13231-1 standard.
For the cant, the reason of the defect is different from the reason highlighted for the track V1. So, we will
study one of these sections in order to define the reasons of this non-compliance.
95% of compliance
99% of compliance
TIP5-CT-2006-031312 Page 82 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
c.v-2 Non-compliance analysis
c.v-2.1 Section [10,2: 10,4[:
Figure 5.1.81: Cant defects in section [10,2: 10,4[
We note that the applied cant is variable in the curve, which generates a cant defect.
TIP5-CT-2006-031312 Page 83 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(vi) Horizontal alignment
The following figure presents the standard deviation for the horizontal alignment per 200m section:
Figure 5.1.82: Standard deviation for horizontal alignment per 200m section V2
c.vi-1 General observations
We note that, on the overall length of the track V1, the mean standard deviation is 1.6mm. Thus, the
horizontal alignment defect range is between 3.2mm et +3.2mm for a 95% compliance and between
4.8mm et +4.8mm for a 99% compliance.
However, considering a 95% compliance, we note that, in several sections, the standard deviation is
higher than the acceptance tolernace of the EN 13231-1 standard.
Remider of the reasons of horizontal alignment defects for track V1
Defect related to post-processing fault.
95% of compliance
99% of compliance
TIP5-CT-2006-031312 Page 84 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(vii) Compliance of the track V2 to the EN 13231-1 standard
For the overall length of the track V2 and for each parameter, the percentage of non-compliant track is
listed in the table below:
Geometric parameter Acceptance tolerances % of non-compliance
Gauge 3mm 5,10
Vertical alignment 6mm 0,29
Cant 3mm 1,45
Twist 4.5mm 4,69
Horizontal alignment 5mm 2,73
Table 5.1.19: Percentage of non-compliant track V2
In order to highlight where these non-compliance are located, we will analyse these parameters in
relation to the curve radius. The table presented hereafter shows the results:
25 < R s 150 150 < R s 500 500 < R s 2000 R >2000 and AD
Geometric parameter Tolerances
% of non-
compliance
% of non-
compliance
% of non-
compliance
% of non-
compliance
Gauge 3mm 4,71* 3,50 4,52 1,74
Vertical alignment 6mm 0,00* 0,14 0,16 0,08
Cant 3mm 4,68* 1,25 0,83 0,43
Twist 4.5mm 0,00* 10,05 0,94 0,73
Horizontal alignment 5mm 9,52* 0,36 1,26 0,11
* Due to a computer issue, this analysis has been done only on 0.9 of the 4km of track having a curve radius range between 25m and 150m.
Table 5.1.20: Percentage of non-compliant track per parameter and curve radius V2
The gauge is the most sensitive parameter for concrete slab tracks (directly related to the wear
phenomenon in small radius curves), we will compare the repartition graph in relation to the curve
radius.
TIP5-CT-2006-031312 Page 85 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Figure 5.1.83: Comparison of gauge repartition graph in relation to the curve radius V2
(viii) Conclusion
The percentage of non-compliant track for vertical alignment, cant, twist and horizontal alignment is
lower than 5%. Thus, considering 95% compliance, we conclude that the track V2 of the ML-3 line is
compliant to the EN 13231-1 standard acceptance tolerances for these track parameters.
For the gauge, the percentage of non-compliant track is a bit higher than 5%. A part of this percentage can
be related to the wear phenomenon in small radius curves and the rest may be related to track installation
issues.
For the other parameters, the percentage of non-compliance is either due to a track installation issue or
due to a post-processing defect. So, in order to refine the compliance analysis, the track design for each
parameter must be an input data for post-processing.
According to the table 6, we can note that, in general, the percentage of non-compliant track increases
when the radius curve is decreasing, except for horizontal alignment and gauge in curve radius from
500m to 2000m. In addition, we note that for the horizontal alignment and cant, the percentage of non-
compliant track is mainly focused on the curve radius range between 25 and 150m and for the twist, it is
focused on the curve radius range between 150m and 500m.
According to the comparison of the gauge repartition graphs in relation to the curve radius, we can note
that, for straight alignment and radius curves higher than 2000m, the mean gauge is around 1435mm. We
can also note that this mean gauge increases when the curve radius is decreasing up to 1436mm for the
curve radius range between 25m and 150m.
TIP5-CT-2006-031312 Page 86 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.4.4. Roughness Measurement
(a) THE MEASUREMENT TROLLEY
The CAT trolley (Corrugation Analysis Trolley) is made up of a measuring head which is adjusted to run
in the middle of one rail running band, a stabilising pole, a pushing pole and a laptop, which records the
signal. A measurement is recorded every 1mm.
Figure 5.1.84: CAT trolley
(b) THE MEASUREMENTS
The measurements were carried out between the 10
th
and the 12
th
of May 2010 on 10 sites of Madrid
tramway line 3. These sites are usually 100m long and were selected as representative samples of the line.
For each site, both rails were measured twice in the middle of the running band.
In this report, right and left rails are considered as in the Figure below:
Figure 5.1.85: Right-left rail definition
The results show a good repeatability.
Travel direction
right
right
left
left
Travel direction
TIP5-CT-2006-031312 Page 87 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
The following table gives an overview of the measured sites.
Site track kp Alignement Type Observation
1 1 13+200-
13+300
Straight classic Slope
2 1 11+200-
11+300
Straight comfort Acceleration slope
3 1 11+100-
11+200
Straight comfort Slope
4 1 10+900-
11+000
Straight comfort Slope
5 1 10+820-
10-852
Curve classic
6 1 3+300-
3+400
Straight classic
7 2 1+620-
1+720
Straight classic
8 2 1+552-
1+592
S-curve classic Acceleration
9 2 1+215-
1+290
Straight Classic + manta No lateral mat
10 2 0+780-
0+880
Straight Classic + manta No lateral mat
MLOreports corrugation in this area
Table 5.1.21: Measurement sites
Pictures of the 10 sites are shown below.
Sites 1 to 4 present an important slope.
Paving is generally deactivated concrete or impressed concrete, which are clean surfaced. But in a few
sites, coating are paving stones, which are sometimes damaged by the expansion or dirt. These types of
coating provokes dust to be on the rails and thus premature wear
Site 1
TIP5-CT-2006-031312 Page 88 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Site 2 & 3
Site 4
Situated station
Site 5
TIP5-CT-2006-031312 Page 89 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Site 6
Site 7
Site 8
Site 9
TIP5-CT-2006-031312 Page 90 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Site 10
(c) RESULTS
The data analysis where carried out in line with both standards ISO 3095 and ISO 13231-3. In the present
chapter, the different results obtained are presented.
(d) ISO3095
(i) Roughness levels
Roughness levels are calculated in 1/3-octave bands as defined in standard ISO 3095. The roughness level
represented in the graphs below is the mean value of both rails roughness measured twice.
Target ISO 3095 level for reference tracks is indicated in red and the thick blue lines represents the range
within which the roughness of Montpellier tramway network is situated. Great variations of roughness
are usually met along one line.
For classic and manta track types, roughness variation according to measurement site is small. For
wavelengths smaller than 200mm, the roughness is within 5dB of ISO 3095 limit for reference track.
Extra high performance jacket track type sections measured appear to have a much higher roughness.
This can be explained by the coating, which is dirt on site 4. On sites 2 and 3, the pavement coating is
worn. The graph below shows that site 4 rails roughness is much higher than the other two.
The sections measured here are on descending slopes, and the use of sand for breaking and / or a higher
velocity can cause premature wear.
For long wavelengths, roughness is higher than on other tramway networks.
Moreover, corrugation with a wavelength 50-63mm appears on site 2.
TIP5-CT-2006-031312 Page 91 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Figure 5.1.87: Classic and floating slab track types roughness spectra
Figure 5.1.88: Comfort track type roughness spectra
TIP5-CT-2006-031312 Page 92 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii) Conclusions
The track roughness measured on Madrid network is particularly good in comparison with what could
be found on some other networks.
The reason for this good track roughness is somewhat difficult to find even after analysis of measurement
collected.
Is it related to:
Madrid segregated track network?
Grinding performed from March to November 2008?
Continuously supported rails?
Low traffic?
It is difficult to draw a conclusion at the moment.
TIP5-CT-2006-031312 Page 93 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.4.5. Pass-by Measurement: Noise and Vibration
(a) SET-UP
The noise and vibration levels of vehicles passing by were measured on 19 May 2010. This was done for
the three different systems/sections.
Accelerometers are placed on a straight line, perpendicular to the rail at distances 3 m (V1), 4 m (V2), 6 m
(V3) and 8 m (V4) from the outer rail.
Microphones are placed at 7.5 m from centre track at 1.2m (N1) and 3.5 m (N2) above top of rail. Vehicles
drive at a constant speed of 30 km/h for all sections. The vehicle speed is verified using a speed radar.
Following paragraphs show pictures of the measurement setups for each site (system).
(b) SITE S1 CLASSIC + MANTA
Figure 5.1.89: Picture of Site S1 Classic + Manta
TIP5-CT-2006-031312 Page 94 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(i) Site S2 Classic
Figure 5.1.90: Picture of Site S3 Classic
(ii) Site S3 Comfort
Figure 5.1.91: Picture of Site S3 Comfort
TIP5-CT-2006-031312 Page 95 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(c) RESULTS OF NOISE MEASUREMENTS
All detailed measurement results can be found in appendix, showing the LMAX and LEQ spectra and the
overall levels in function of the time for each pass-by measurement.
The measurements are summarized in tables Table 5.1.21. The A-weighted noise levels are about 3dB(A)
higher for the system Classic + Manta than for Classic and Comfort. The noise levels for those last
two are similar.
Site System Direction N1
[dB(A)]
N2
[dB(A)]
N1
[dB(A)]
N2
[dB(A)]
S1: 0+820 Classic + Manta Colonia Jardin 73.4 72.5 70.3 69.1
S2: 1+760 Classic Colonia Jardin 70.4 69.8 67.9 67.1
S3: 13+060 Comfort Colonia Jardin 69.8 70.2 67.2 67.4
Table 5.1.21: Overview
(d) RESULTS OF VIBRATION MEASUREMENTS
All detailed measurement results can be found in appendix, showing the LMAX and LEQ spectra and the
overall levels in function of the time for each pass-by measurement.
The measurements are summarized in table
Site
System Direction LEQ V2 [dB(re.1e-9m/s)]
V1
V2
V3 V4
S1: 0+820 Classic + Manta Colonia Jardin 92.9 90.7 88.2 85.5
S2: 1+760 Classic Colonia Jardin 90.9 87.5 85.3 84.3
S3: 13+060 Comfort Colonia Jardin 89.7 86.5 82.3 83.8
Table 5.1.22. he global LEQ vibration level for Classic + Manta is 1.2 to 3.2 dBv higher than for Classic. The
global LEQ vibration level for Comfort is 0.5 to 3.0 dBv lower than for Classic.
Site System Direction LMAX [dB(re.1e-9m/s)]
V1 V2 V3 V4
S1: 0+820 Classic + Manta Colonia Jardin 97.4 97.1 95.0 92.1
S2: 1+760 Classic Colonia Jardin 96.0 92.4 90.5 89.9
S3: 13+060 Comfort Colonia Jardin 92.5 89.9 86.0 88.3
Site System Direction LEQ V2 [dB(re.1e-9m/s)]
V1
V2
V3 V4
S1: 0+820 Classic + Manta Colonia Jardin 92.9 90.7 88.2 85.5
S2: 1+760 Classic Colonia Jardin 90.9 87.5 85.3 84.3
S3: 13+060 Comfort Colonia Jardin 89.7 86.5 82.3 83.8
Table 5.1.22: Overview
TIP5-CT-2006-031312 Page 96 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(e) CONCLUSION
The approach pertaining to vibration mitigation has to consider the whole vibration transmission path
and to be adapted to the vehicle that will be operated.
And the vibration performance of a transportation system must be defined in relation to its compliance
with a level of vibration measured at building locations and not by empirical rules based on outdated
data.
The measurements made on many other sites for several tram designs and manufacturers are correlating
what could be observed in Madrid:
A very high performance system (floating slab) is rarely required,
A continuously supported rail system is sufficient in many cases.
TIP5-CT-2006-031312 Page 97 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.5 Metro track form
5.1.5.1. The Mathematical Model
In the following paragraph the main characteristics of the numerical model of train-track interaction as
well as of the wheel-rail contact model are briefly described.
(A) TRAIN-TRACK INTERACTIONMODEL
An efficient multibody modelling approach for the study of the running behaviour of railway vehicles
was developed in recent years by a research group established at the Department of Mechanical
Engineering, Politecnico di Milano [1], [2], [3].
The mathematical model is based on a multi-body, large displacement schematisation of the trainset,
allowing to analyse the non-stationary behaviour considering tangent track running and curve
negotiation.
The system under study is subdivided into elementary units of the following types:
Carbody, modelled as a single rigid body;
Bogie assembly, modelled as a rigid bogie frame connected by primary suspensions to two flexible
wheelsets;
Other bodies, e.g. motors, converters, assumed to displace rigidly and being attached either to a
carbody or to a bogie frame.
Elementary units are connected to each other by means of elastic and damping elements (linear and
nonlinear) reproducing secondary suspensions, links between carbodies, and connections to other bodies
(e.g. elastic motor suspension). By combining the above listed elementary units, any trainset architecture
may be derived, such as rail vehicles formed by two bogies and one carbody, or more complex
configurations formed by several articulated carbodies.
Each rigid body is assigned with 5 degrees of freedom, the forward speed of body centre of mass being
set to a constant value V, whereas for each flexible wheelset the movement with respect to the moving
reference is defined as the linear combination of the unconstrained wheelset eigenvectors. For the study
of vehicle stability and of ride safety, the model can be limited to the five modes corresponding to the
rigid motions of the unconstrained wheelset (again considering a forward motion at constant speed),
whereas in the study of high-frequency interaction of the train with the track (with applications e.g. to
turnouts transit) bending and torsion modes as well as some modes corresponding to local deformation
of the wheels are typically included in the analysis.
To represent kinematical effects associated with curve negotiation, the motion of each elementary unit is
described with respect to a moving reference system travelling at constant speed along the track
centreline, with Z axis tangent to the track centreline and X axis orthogonal to the rail level. By assuming
small displacements relative to the moving reference of each module in the trainset, the equations of
motion are linearised with respect to kinematical nonlinear effects only, and take the form:
TIP5-CT-2006-031312 Page 98 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
| | | | | | ( ) ( ) ) , , , , , ( ) , ( , , t V X X X X F X X F t V F t V F X K X C X M
T T v v cv v v nl i e v v v v v v

+ + + = + +
(1)
where | |
v
M , | |
v
C and | |
v
K are the mass, damping and stiffness matrices of the trainset,
v
X is the
vector of trainset coordinates, ( ) t V F
e
, is the vector of generalised forces produced in the secondary
suspensions and carbody links by the different motions of the moving references associated with the
modules connected by the suspension, ( ) t V F
i
, is the vector of inertial forces due to the non-inertial
motion of the moving references,
nl
F is the vector of nonlinear forces due to nonlinear elements in the
suspensions and
cv
F is the vector of generalized forces due to wheel-rail contact, depending upon the
motion of the trainset
v v
X X

, and upon additional coordinates
t
X representing the track motion and
their time derivatives
t
X

. The dynamic behaviour of the track being negotiated by the train is

represented by the additional equation:
| | | | | | ) , , , , , ( t V X X X X F X K X C X M
t t v v ct t t t t t t

= + +
(2)
where | |
t
M , | |
t
C and | |
t
K are the track mass, damping and stiffness matrices, defined by a linear finite
element schematisation, and
ct
F is the vector of generalized nodal forces acting on the track due to
wheel rail contact forces.
Equations (1) and (2) take the form of two coupled sets of nonlinear differential equations, with the
coupling term being provided by wheel-rail contact forces.
Figure 5.1.92: Train-track dynamic interaction model
TIP5-CT-2006-031312 Page 99 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(B) WHEEL-RAIL CONTACT MODEL
Being the vehicle and track subsystems considered separately, the contact forces exchanged at wheel-rail
interface are expressed as a function of both subsystems motion. This is why the independent coordinates
Xt and Xv appear both in the track and in the vehicle generalized forces, as coupling terms between the
two sets of equations.
At each integration step, the procedure adopted for contact forces computation follows the scheme
reported below:
1. a first attempt solution for the independent variables Xt and Xv is predicted;
2. on the basis of the shape functions of the track finite element model, the track displacement at each
wheel-rail contact point is calculated, together with its time derivative;
3. using modal superposition or rigid body kinematics relations, the vehicle displacement at each
wheel-rail contact point is calculated, together with its time derivative;
4. starting from the displacements mentioned in items 2 and 3 and their time derivatives, the normal
(Nj) and tangential (FTj, FLj) contact forces acting on the j-th wheel-rail couple are evaluated (Figure 2);
5. using again the shape functions of the track finite element model and the vehicles modal shapes, the
contribution of the forces Nj, FTj, FLj to the generalised forces Fcv and Fct acting on the vehicle and on
the track degrees of freedom is evaluated; steps from 2) to 5) are repeated for all wheels;
6. finally, once that the generalised forces Fcv and Fct are determined, the equations (1) and (2) are
solved, and a better approximation of the solution (Xt and Xv) is achieved.
As far as point 4) is concerned, the computation of the normal contact forces Nj is based on a multi-
hertzian model (Bruni et al. 1999), while the tangential forces FTj and FLj are obtained according to the
Shen, Hedrick & Elkins formulation (1983).
The wheel-rail contact model (Bruni et al. 1999) is based on a preliminary geometrical analysis, which can
be carried out both on new and on measured wheel/rail worn profiles. Contact parameters such as the
local radii of curvature and the contact angle are reported in table form, as functions of wheel-rail lateral
displacement. The geometrical analysis also allows to determine the number of the potential contact
points for a given wheel-rail relative position.
The sets of equations (1) an (2) are numerically integrated using a modified Newmark method. At each
integration step, convergence on the wheel-rail contact forces is reached iteratively. In this way, the
dynamic response of each subsystem is automatically dependent on the motion of the other part of the
system. Moreover, the forcing effect associated with the rail roughness is accounted for, in terms of rail
displacements superimposed to the ones calculated, at the generic integration step, as a function of the
track nodal coordinates. For given irregularity profile and train velocity, the former displacements are
only a function of time.
As a result of the integration procedure, the time histories of the following quantities can be obtained:
Track displacements/accelerations;
Vehicle displacements/accelerations;
Forces transmitted from the vehicle to the track.
TIP5-CT-2006-031312 Page 100 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
rail and wheel
profiles
contact geometrical
parameters
geometrical
analysis
elastic deformation in normal
direction (penetration)
tangential & longitudinal
creepages
generalized
contact forces
tangential & longitudinal forces
(Shen-Hedrick-Elkins model)
normal forces
(multi-hertzian model)

R
rail and wheel
profiles
contact geometrical
parameters
geometrical
analysis
elastic deformation in normal
direction (penetration)
tangential & longitudinal
creepages
generalized
contact forces
tangential & longitudinal forces
(Shen-Hedrick-Elkins model)
normal forces
(multi-hertzian model)

R
Figure 5.1.93: Wheel-rail contact model
TIP5-CT-2006-031312 Page 101 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.5.2. Metro Madrid
(a) VEHICLE ANDTRACK CHARACTERISTICS
Figure 5.1. shows a sketch of the implemented model of the Metro Madrid coach. The main vehicle
characteristics are:
Maximum axle load = 154kN;
Tare axle load = 77kN;
Wheelbase = 2.2 m;
Pivot pitch = 11.1m.
Figure 5.1.94: Sketch of the implemented model of the Metro Madrid coach
In all the simulations presented hereafter, the vehicle has been assumed in full load condition (axle load =
154kN). The elastic and damping characteristics of both the primary and secondary suspensions are taken
from a metro vehicle with similar characteristics.
Figure 5.1.93 shows the considered wheel irregularity (multiplied by a factor of 1000). Note that the shape
of such irregularity changes form wheel to wheel, but has the same wavelengths and wave amplitudes.
TIP5-CT-2006-031312 Page 102 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
0.1
0.2
0.3
0.4
0.5
30
210
60
240
90
270
120
300
150
330
180 0
Figure 5.1.95: Wheel irregularity (multiplied by 1000): right wheel (blue), left wheel (red).
Track irregularity input has been defined according to ORE B176 PSD function (small amplitude defects).
Figure 5.1.94 shows the PSD function of the track vertical profile, while figure 5.1.95 shows the track
vertical irregularity profile, reconstructed on the basis of a random phase generation. Note that, although
not shown here, the ORE irregularity provided in input to the train-track interaction simulation includes
also cross-level and alignment irregularities.
As far as the wheel and rail profiles are concerned, the ORE S1002 and the UIC54 rail were considered.
(i) Track Model
For the purpose of this research, a numerical model of the embedded-rail track system has been
developed (Figure 5.1.) and implemented into PoliMis train-track dynamic simulation code. The linear
visco-elastic characteristics of the REMS elastic bed take place between rails and tunnel (both schematised
as equivalent Euler-Bernoulli beams). The bending stiffness and the mass per unit length of the rail can be
easily determined form the rail producer catalogue, the properties of the equivalent beam representing
the tunnel and those of the visco-elastic bed representing the ground were identified from impulsive tests
carried out some years ago in Milano underground. The visco-elastic bed representing the ground has
been introduced to determine the transmissibility (i.e. the capability of filtering out vibrations induced by
train passage) of the considered REMS tracks.
Figure 5.1.96. Implemented model of the REMS track system
TIP5-CT-2006-031312 Page 103 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Stiffness characteristics of the embedded rail system were originally calculated by means of 2D FEM
analyses. As an example Figure 5.1. shows the computed static response of the system (displacements
and Von Mises stresses in the rubber jacket) for the comfort sample. This study was useful to design the
three samples as a function of the desired train-track interaction. Once stiffness characteristics were
determined, samples were manufactured by CDM and subsequently tested in PoliMi labs.
Figure 5.1.97. Embedded system static Von Mises stresses, Comfort sample.
For the three samples, in order to identify the real REMS visco-elastic bed characteristics, a series of tests
were performed (Figure 5.1.95) in two different configurations (Figure 5.1.10 and
Figure 5.1.11) useful to simulate typical loads for straight and curved track. Thanks to these tests it was
possible to determinate vertical / lateral equivalent stiffness and damping of the visco-elastic bed for
each of the three REMS samples. Also fatigue and aging tests were performed in order to evaluate
durability and possible variations of the identified parameters (see deliverables D1.1 and D3.1 for more
details about the experimental test methodology and results).
TIP5-CT-2006-031312 Page 104 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Figure 5.1.10. 0-degrees test configuration, straight
track.
Figure 5.1.98. Picture of the PoliMi test bench in the C4
laboratory, Milan
Figure 5.1.11. 26-degrees test configuration, curve
track.
For each sample both static and dynamic characterisation has been performed. As an example in Figure
5.1. the vertical stiffness of the Comfort sample is reported. During dynamic tests the sample has been
forced with different actuation frequency and pre-load. The resulting stiffness is therefore reported as a
function of these two parameters.
0 5 10 15
0
10
20
30
40
50
60
70
80
90
100
[Hz]
[
M
N
/
m
/
m
]
COMFORT Vertical stiffness per unit length Load amplitude: 10% of Q
0
Sample inclination: 0 deg Before fatigue test
Preload: 10% of Q
0
Preload: 30% of Q
0
Preload: 50% of Q
0
Preload: 70% of Q
0
Preload: 90% of Q
0
Figure 5.1.99. Vertical stiffness of the Comfort sample, function of the preload and of the actuation frequency.
TIP5-CT-2006-031312 Page 105 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.5.3. Numerical Results
The collected data (see experimental evaluation of the REMS visco-elastic bed etc.) both for train and
track characteristics, allow PoliMis simulation code to provide the dynamic train-track response due to a
known track geometry and train speed for the three types of REMS tracks considered in the present
research, i.e. compact, classic and comfort tracks. Both wheel and rail irregularity input data were
considered in all the simulations. While measured irregularity patterns were considered for the wheel,
track irregularity input was generated according to the PSD functions defined in the ORE B176 Standard.
As an example, Figure 5.1. shows the numerically simulated time histories of the vertical rail and tunnel
accelerations during tangent track running at 110km/h. Results are reported for the three REMS tracks
and for the traditional direct fixation track. It can be clearly seen that the REMS track is able to reduce
both rail and tunnel accelerations, in terms of amplitudes and of duration, since the longitudinal
transmission of the vibration is better damped.
(a) Comfort
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
-80
-60
-40
-20
0
20
40
60
80
s
m
/
s
2
rail
tunnel
(b) Classic
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
-80
-60
-40
-20
0
20
40
60
80
s
m
/
s
2
rail
tunnel
(c) Compact
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
-80
-60
-40
-20
0
20
40
60
80
s
m
/
s
2
rail
tunnel
(d) Direct fixation
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
-100
-50
0
50
100
150
[s]
[
m
/
s
2
]
vertical acceleration
rail
tunnel
Figure 5.1.100. Numerical simulations for the estimation of vibration mitigation performance (REMS and direct
fixation). Metro Madrid train running in tangent track at 110km/h (full load). Simulations output:
rail and tunnel acceleration time histories.
TIP5-CT-2006-031312 Page 106 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
In the following, as an example, other typical results for assessing the dynamic behaviour of the track are
shown: vertical rail displacement and gauge widening (Figure 5.1.12 and Figure 5.1.) during curved track
running. All results are plotted when the first axle of Metro Madrid vehicle has covered a significant
length in full curve.
As expected, the vertical displacement of the rail is higher for the REMS tracks than for the direct fixation
track. However, even in the worst case (comfort track), the vertical displacement of the REMS track is
smaller than 3 mm. The gauge widening is almost equal for both the REMS and the direct fixation track,
due to similar lateral deformability. An absolute value of about 1.2 mm can be observed for the comfort
track on a 300 m radius curve for the fully loaded vehicle.
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
Vertical rail deflection - R = 300 m, Cant = 150 mm, V = 80 km/h, Full load (axle load =154 kN)
[s]
[
m
m
]
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
0
0.2
0.4
0.6
0.8
1
Gauge widening - R = 300 m, Cant = 150 mm, V = 80 km/h, Full load (axle load =154 kN)
[s]
[
m
m
]
20m out of transition
inner rail - 20m out of transition
outer rail - 20m out of transition
Figure 5.1.121. Vertical rail displacement for the Comfort track (full load) and gauge widening.
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
Vertical rail deflection - R = 300 m, Cant = 150 mm, V = 80 km/h, Full load (axle load =154 kN)
[s]
[
m
m
]
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
0
0.2
0.4
0.6
0.8
1
Gauge widening - R = 300 m, Cant = 150 mm, V = 80 km/h, Full load (axle load =154 kN)
[s]
[
m
m
]
20m out of transition
inner rail - 20m out of transition
outer rail - 20m out of transition
Figure 5.1.102. Vertical rail displacement for the Direct Fixation track (full load) and gauge widening.
TIP5-CT-2006-031312 Page 107 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
To assess track transmissibility, the power spectral densities of both the rail and the tunnel vertical
accelerations were evaluated (Figure 5.1.). Thus, the transfer function of the tunnel vertical acceleration
with respect to the rail vertical acceleration was computed showing the good filtering properties of all
three REMS tracks (Figure 5.1.13). Note that the comfort track is able to completely filtering out the
resonance peak at approx. 25 Hz.
10
1
10
2
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
10
1
[Hz]
P
S
D
[
(
m
/
s
2
)
2
/
H
z
]
Power spectral density of tunnel & rail vertical accelerations
Classic Rail
Classic Tunnel
Classic+10% Rail
Classic+10% Tunnel
Classic+20% Rail
Classic+20% Tunnel
Comfort Rail
Comfort Tunnel
Comfort+10% Rail
Comfort+10% Tunnel
Comfort+20% Rail
Comfort+20% Tunnel
Compact Rail
Compact Tunnel
Compact+10% Rail
Compact+10% Tunnel
Compact+20% Rail
Compact+20% Tunnel
Figure 5.1.103. Numerically evaluated power spectral densities of both the rail and the tunnel vertical accelerations
10
1
10
2
10
-4
10
-3
10
-2
10
-1
10
0
[Hz]
T
F
[
-
]
Transfer function tunnel vertical acceleration/rail vertical acceleration
Classic Rail
Classic+10% Rail
Classic+20% Rail
Comfort Rail
Comfort+10% Rail
Comfort+20% Rail
Compact Rail
Compact+10% Rail
Compact+20% Rail
Figure 5.1.134. Numerically evaluated transfer function of the tunnel vertical accelerations wrt the rail vertical
accelerations
TIP5-CT-2006-031312 Page 108 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Figure 5.1. shows the maximum value of wheel-rail contact force inclination angle for the compact track
(a), for the classic track (b) and for the comfort track (c) with zero preload and both tare and full loaded
vehicles. It can be seen that the highest inclination angle is equal to 21 deg and is obtained for the comfort
track for the 300 m radius curve. Note that the results shown in figure 10 are obtained by using a one
metre moving average to assess the mean behaviour during curve negotiation thus filtering out spikes
associates to track and / or wheel irregularity. The maximum inclination angle is used for determining
the fatigue behaviour of the tracks.
0
5
10
15
20
[
d
e
g
]
Maximumvalue of wheel-rail force inclination angle
C
o
m
p
a
c
t
R
=
2
1
0
m
f
u
l
ly
lo
a
d
e
d
C
o
m
p
a
c
t
R
=
2
1
0
m
t
a
r
e
lo
a
d
C
o
m
p
a
c
t
R
=
3
0
0
m
f
u
l
ly
lo
a
d
e
d
C
o
m
p
a
c
t
R
=
3
0
0
m
t
a
r
e
lo
a
d
C
o
m
p
a
c
t
R
=
5
0
0
m
f
u
l
ly
lo
a
d
e
d
C
o
m
p
a
c
t
R
=
5
0
0
m
t
a
r
e
lo
a
d
C
o
m
p
a
c
t
R
=
1
0
0
0
m
f
u
ll
y
l
o
a
d
e
d
C
o
m
p
a
c
t
R
=
1
0
0
0
m
t
a
r
e
l
o
a
d
C
o
m
p
a
c
t
s
t
r
a
i
g
h
t
t
r
a
c
k
f
u
ll
y
l
o
a
d
e
d
C
o
m
p
a
c
t
s
t
r
a
i
g
h
t
t
r
a
c
k
t
a
r
e
l
o
a
d
0
5
10
15
20
[
d
e
g
]
Maximumvalue of wheel-rail force inclination angle
C
l
a
s
s
i
c
R
=
2
1
0
m
f
u
ll
y
l
o
a
d
e
d
C
l
a
s
s
i
c
R
=
2
1
0
m
t
a
r
e
l
o
a
d
C
l
a
s
s
i
c
R
=
3
0
0
m
f
u
ll
y
l
o
a
d
e
d
C
l
a
s
s
i
c
R
=
3
0
0
m
t
a
r
e
l
o
a
d
C
l
a
s
s
i
c
R
=
5
0
0
m
f
u
ll
y
l
o
a
d
e
d
C
l
a
s
s
i
c
R
=
5
0
0
m
t
a
r
e
l
o
a
d
C
l
a
s
s
i
c
R
=
1
0
0
0
m
f
u
l
l
y
l
o
a
d
e
d
C
l
a
s
s
i
c
R
=
1
0
0
0
m
t
a
r
e
l
o
a
d
C
l
a
s
s
i
c
s
t
r
a
ig
h
t
t
r
a
c
k
f
u
l
l
y
l
o
a
d
e
d
C
l
a
s
s
i
c
s
t
r
a
ig
h
t
t
r
a
c
k
t
a
r
e
l
o
a
d
0
5
10
15
20
25
[
d
e
g
]
Maximumvalue of wheel-rail force inclination angle
C
o
m
f
o
r
t
R
=
2
1
0
m
f
u
l
l
y
l
o
a
d
e
d
C
o
m
f
o
r
t
R
=
2
1
0
m
t
a
r
e
l
o
a
d
C
o
m
f
o
r
t
R
=
3
0
0
m
f
u
l
l
y
l
o
a
d
e
d
C
o
m
f
o
r
t
R
=
3
0
0
m
t
a
r
e
l
o
a
d
C
o
m
f
o
r
t
R
=
5
0
0
m
f
u
l
l
y
l
o
a
d
e
d
C
o
m
f
o
r
t
R
=
5
0
0
m
t
a
r
e
l
o
a
d
C
o
m
f
o
r
t
R
=
1
0
0
0
m
f
u
l
ly
l
o
a
d
e
d
C
o
m
f
o
r
t
R
=
1
0
0
0
m
t
a
r
e
lo
a
d
C
o
m
f
o
r
t
s
t
r
a
i
g
h
t
t
r
a
c
k
f
u
l
ly
l
o
a
d
e
d
C
o
m
f
o
r
t
s
t
r
a
i
g
h
t
t
r
a
c
k
t
a
r
e
lo
a
d
Figure 5.1.105. Maximum value of wheel-rail contact force inclination angle for the compact track (a), for the classic
track (b) and for the comfort track (c) with zero preload and both tare and full loaded vehicles (1m
moving average)
Another important parameter to assess the dynamic behaviour of REMS tracks during curve negotiation
is the maximum value of the outer wheel lateral contact force component (Figure 5.1.). The same one
metre moving average filter has been used. It can be seen that the highest lateral contact force component
is obtained for the fully loaded vehicle running along the 300 m radius curve of both the compact or
comfort tracks.
0
5
10
15
20
25
30
35
[
k
N
]
Maximumvalue of outer wheel lateral force component
C
o
m
p
a
c
t
R
=
2
1
0
m
f
u
l
ly
lo
a
d
e
d
C
o
m
p
a
c
t
R
=
2
1
0
m
t
a
r
e
l
o
a
d
C
o
m
p
a
c
t
R
=
3
0
0
m
f
u
l
ly
lo
a
d
e
d
C
o
m
p
a
c
t
R
=
3
0
0
m
t
a
r
e
l
o
a
d
C
o
m
p
a
c
t
R
=
5
0
0
m
f
u
l
ly
lo
a
d
e
d
C
o
m
p
a
c
t
R
=
5
0
0
m
t
a
r
e
l
o
a
d
C
o
m
p
a
c
t
R
=
1
0
0
0
m
f
u
ll
y
l
o
a
d
e
d
C
o
m
p
a
c
t
R
=
1
0
0
0
m
t
a
r
e
l
o
a
d
C
o
m
p
a
c
t
s
t
r
a
ig
h
t
t
r
a
c
k
f
u
ll
y
l
o
a
d
e
d
C
o
m
p
a
c
t
s
t
r
a
ig
h
t
t
r
a
c
k
t
a
r
e
l
o
a
d
0
5
10
15
20
25
30
35
[
k
N
]
Maximumvalue of outer wheel lateral force component
C
la
s
s
i
c
R
=
2
1
0
m
f
u
l
l
y
l
o
a
d
e
d
C
la
s
s
i
c
R
=
2
1
0
m
t
a
r
e
l
o
a
d
C
la
s
s
i
c
R
=
3
0
0
m
f
u
l
l
y
l
o
a
d
e
d
C
la
s
s
i
c
R
=
3
0
0
m
t
a
r
e
l
o
a
d
C
la
s
s
i
c
R
=
5
0
0
m
f
u
l
l
y
l
o
a
d
e
d
C
la
s
s
i
c
R
=
5
0
0
m
t
a
r
e
l
o
a
d
C
la
s
s
i
c
R
=
1
0
0
0
m
f
u
l
l
y
l
o
a
d
e
d
C
la
s
s
i
c
R
=
1
0
0
0
m
t
a
r
e
lo
a
d
C
la
s
s
i
c
s
t
r
a
i
g
h
t
t
r
a
c
k
f
u
l
ly
l
o
a
d
e
d
C
la
s
s
i
c
s
t
r
a
i
g
h
t
t
r
a
c
k
t
a
r
e
lo
a
d
0
5
10
15
20
25
30
35
[
k
N
]
Maximumvalue of outer wheel lateral force component
C
o
m
f
o
r
t
R
=
2
1
0
m
f
u
l
l
y
l
o
a
d
e
d
C
o
m
f
o
r
t
R
=
2
1
0
m
t
a
r
e
lo
a
d
C
o
m
f
o
r
t
R
=
3
0
0
m
f
u
l
l
y
l
o
a
d
e
d
C
o
m
f
o
r
t
R
=
3
0
0
m
t
a
r
e
lo
a
d
C
o
m
f
o
r
t
R
=
5
0
0
m
f
u
l
l
y
l
o
a
d
e
d
C
o
m
f
o
r
t
R
=
5
0
0
m
t
a
r
e
lo
a
d
C
o
m
f
o
r
t
R
=
1
0
0
0
m
f
u
l
ly
lo
a
d
e
d
C
o
m
f
o
r
t
R
=
1
0
0
0
m
t
a
r
e
lo
a
d
C
o
m
f
o
r
t
s
t
r
a
i
g
h
t
t
r
a
c
k
f
u
l
ly
lo
a
d
e
d
C
o
m
f
o
r
t
s
t
r
a
i
g
h
t
t
r
a
c
k
t
a
r
e
lo
a
d
Figure 5.1.106. Maximum value of the outer wheel lateral contact force component for the compact track (a), for the
classic track (b) and for the comfort track (c) with zero preload and both tare and full loaded
vehicles (1m moving average)
(a) (b) (c)
(a) (b) (c)
TIP5-CT-2006-031312 Page 109 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.5.4. References
[1] G. Diana, S. Bruni, A. Collina, R. Vigano, G. Triani and R. Di Bianco, ADTRES: A Software for
railway analysis, Proceedings of WCRR 97, Florence, Italy, Vol. 3, pp. 444-447, 1997
[2] S. Bruni and A. Collina, Modelling the Viscoelastic Behaviour of Elastomeric Components: An
Application to the Simulation of Train-Track Interaction, Vehicle System Dynamics, Vol. 34,
pp. 283-301, 2000
[3] F. Cheli, G. Diana and F. Resta, Numerical Model of a Tilting Body Railway Vehicle Compared
With Rig and on Track Tests, Vehicle System Dynamics, Vol. 35, N. 6, pp. 417-442, 2001
[4] F. Braghin, S. Bruni, A. Collina and R. Corradi, Numerical Simulation of Train-Track
Interaction for the Dynamic Performance Estimation of an Innovative Metro Embedded-Rail
Track System, Proceedings of the 7th European Conferenc on Structural Dynamics, Southampton,
UK, 7-9 July 2008
TIP5-CT-2006-031312 Page 110 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.6 General conclusion
5.1.6.1. Conclusions on Tramway Track form
A wide range of subjects was dealt with in the present document, all of them related to the forces applied
to the track and the resulting stresses in rail, and leading to conclusions on track geometry, track
roughness and vibration mitigation.
Complete models including track and vehicles were setup allowing for simulations taking into account
the actual conditions of the track.
(a) TRACK GEOMETRY
The track geometry was studied in relation with its mechanical response and the resulting forces and
stresses undergone. Following the measurements in Madrid of track for gauge, vertical alignment, and
cant and horizontal alignment, an approach based on 95% compliance can be applied to conclude that the
track V1 of the ML-3 line is compliant to the EN 13231-1 standard acceptance tolerances for these track
parameters.
The results of the track geometry study can be proposed as a contribution for the development of urban
track standards.
(b) RAIL FATIGUE
The rail fatigue was studied with two practical purposes: to be able to make a commitment on the lifetime
of the rails installed and to optimise the rail replacement frequencies for the track maintenance.
The development of that study was the opportunity to implement the stationary method, which enables
to undertake fast simulation of a high number of cycles.
(c) TRACK ROUGHNESS
The track roughness measured on Madrid network is particularly good in comparison with what could
be found on some other networks.
The reason for this good track roughness is somewhat difficult to find even after analysis of measurement
collected.
It is difficult to draw a conclusion at the moment.
(d) VIBRATIONMITIGATION
The approach pertaining to vibration mitigation has to consider the whole vibration transmission path
and to be adapted to the vehicle that will be operated.
And the vibration performance of a transportation system must be defined in relation to its compliance
with a level of vibration measured at building locations and not by empirical rules based on outdated
data.
The measurements made on many other sites for several tram designs and manufacturers are correlating
what could be observed in Madrid:
A very high performance system (floating slab) is rarely required,
A continuously supported rail system is sufficient in many cases.
TIP5-CT-2006-031312 Page 111 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.1.6.2. Conclusions on Metro Track form
Finally the study of the Madrid metro track form based on the REMS system was finalised following the
calibration and adjustment phases carried out during laboratory testing.
TIP5-CT-2006-031312 Page 112 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.2 FONCTIONAL SPECIFICATIONS FOR TRACK INFRASTRUCTURE
5.2.1 Sensitivity analysis on tramway track parameters by means of numerical simulation
of train-track dynamic interaction
Within WP5.2, a numerical train track interaction model was set-up to carry out a sensitivity analysis on
tramway and metro track design parameters to evaluate their effects on track stability and vibration
transmission. The following parameters were considered:
different track design options (embedded rail, direct fixation, floating slab track) and rail support
stiffness values;
support stiffness variations corresponding to new/degraded track conditions;
axle load variations (tare and full load);
track irregularity corresponding to two different levels of degraded track geometry.
This work consisted of four different sections:
sensitivity analysis for tramway tracks
sensitivity analysis for metro tracks
light rail parameters
evaluation of existing standards
overview of a new embedded track standard under development in the US by APTA/AREMA
5.2.1.1. Introduction
As a part of functional requirements for urban railway tracks, knowing the influence of several
parameters on important factors such as track stability and vibrations. It is impossible to carry out
measurements in repeatable conditions on different tracks with different vehicles at different speeds.
Therefore, it has been decided to carry out this sensitivity analysis numerically with a software acquired
by D2S and APT: Vi-Rail. It is a multi-body software dedicated to railway applications.
With the help of this software, along with a Finite Element Software (MSC.Nastran) and a programming
tool (Matlab), simulations were carried out with following varying parameters:
(a) VEHICLE TYPE
Tramway
o Rail type: EB50T, NP4am, 35G Ri59N, Ri53N;
o Track layout: 30 m radius curved track and straight track;
o Vehicle velocity: 20 km/h in curve and 20, 40, 60 km/h on straight track.
Light Rail
o Rail type: UIC50, UIC54;
o Track layout: 100 m radius curved track and straight track;
o Vehicle velocity: 20, 40 and 60 km/h in curve and 20, 40, 60, 80, 100 km/h on straight track.
TIP5-CT-2006-031312 Page 113 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(b) TRACK TYPE
Standard ballasted track with fixation every 60 and 75 cm;
Track with rail directly fastened to a concrete slab: soft and hard railpad stiffnesss with fixation
every 60 and 75 cm;
Track with continuously supported rail: CDM-Classic and CDM-Comfort.
(c) RAIL IRREGULARITIES
Measured rail irregularities
Step function rail irregularity
The combination of these parameters leads to a large number of simulation. The first section of this report
presents the methodology used to carry out all these simulation. The values used for all the parameters
are presented in the second section while the last chapter discusses the results.
5.2.1.2. Description of the methodology
The aim of this sensitivity analysis is to study the influence of several parameters (type of rail, type of rail
support, type of vehicle, speed ) that one will call the input and two major quantities: the vibrations
under the track and the stresses in the rail generated by a passage of a vehicle. The latter quantities can be
considered as the outputs of the model.
The methodology is based on using several software to calculate the requested quantities. The figure
below is schematic view of the methodology.
Figure 5.2.1 Methodology used in the study
TIP5-CT-2006-031312 Page 114 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(a) CALCULATIONOF THE FORCES AND DISPLACEMENTS IN THE TRACK
The forces applied by the vehicle on the rail and the subsequent displacement are calculated in Vi-Rail
and post-processed in Matlab for presentation and figure generation.
(i) Calculations in Vi-Rail
Vi-Rail is a multi-body simulation software dedicated to railway applications. VI-Rail is built upon
MSC.ADAMS, widely recognized as the worlds leading mechanical system simulation tool. This
software allows the simulation of the passage of a given vehicle on a given track at various speeds and to
calculate the forces applied by this vehicle on the rails.
Moreover, rail irregularities can be modelled as well. Measured rail irregularities can be included in the
model: the subsequent generated track displacement are thus varying over time, vibrations are generated.
Using these irregularities therefore allows us to study vibrations generated by a vehicle passage, as one
would measure vibration levels on a real track.
The figure below shows a schematic view of vehicle-track model in Vi-Rail.
Figure 5.2.2 Schematic view of the model in Vi-Rail
(ii) Calculation in Matlab
The forces and displacements calculated with Vi-Rail are exported in text format and read in Matlab for
post-processing.
Forces are plotted in function of time and the average value is saved in a spreadsheet for later use as an
input in the FE rail model for stresses calculation.
The displacements over time are first transformed into spectra in the frequency domain through a Fast
Fourier Transformation (fft). They are then derived to obtain the vibration velocities and plotted in third-
octave band spectra. These spectra are the final results (output) regarding the vibrations. Conclusions can
be directly drawn from there.
TIP5-CT-2006-031312 Page 115 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(b) CALCULATIONOF STRESSES IN THE RAIL INMSC.NASTRAN
The mean forces saved in the spreadsheet are used as input forces on Finite Element analysis on rail
models in MSC.Nastran. MSC.Nastran is a well-known Finite Element structural analysis software
allowing following types of calculation: linear static, modal analysis, thermal-transfer, acoustics,
frequency-transient-spectral response, non-linear static and dynamic, optimization and aeroelasticity.
A Finite Element Model of a section of the rail on its support is model in Patran (pre and port-processing
software for Nastran) and the forces coming from the spreadsheet, the wheel/rail contact forces, are
applied on the rail head to calculate the stresses in the rail web and in the rail foot. These calculated
stresses are the final results regarding the stress calculation and conclusions can be drawn from there.
TIP5-CT-2006-031312 Page 116 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.2.1.3. Presentation of the models and their parameters
In this section, the vehicle and track models are presented as well as their parameters.
(a) PRESENTATION OF THE VEHICLE MODEL
For the purpose of this study, two types of urban vehicle have been modelled: a light rail vehicle and a
tramway vehicle.
(i) Light rail vehicle
The light rail vehicle model is based on the metro fromthe RATP running on line 8.
It is mainly made of a car body on two bogies.
a.i-1 Vehicle car body
Important parameters for the multi body of the light rail vehicles car body are listed below.
Mass (with passengers) 24.5 *10 kg
Mass moment of inertia in roll 15*10 kgm
Mass moment of inertia in pitch 390*10 kgm
Mass moment of inertia in yaw 390*10 kgm
Vertical location of centre of gravity above railhead 1.8 m
Longitudinal distance between bogie centres 10.0 m
a.i-2 Vehicle bogies (without axles)
Important parameters for the multi body of the light rail vehicles bogies without axles are listed below.
Mass 1800 kg
Mass moment of inertia in roll 760 kgm
Mass moment of inertia in pitch 2.2*10 kgm
Mass moment of inertia in yaw 2.2*10 kgm
Vertical location of centre of gravity above railhead 0.5 m
Longitudinal distance between wheel sets 2.0 m
a.i-3 Vehicle axles
Important parameters for the multi body of the light rail vehicles axles are listed below.
Mass 935 kg
Mass moment of inertia in roll 530 kgm
Mass moment of inertia in pitch 85 kgm
Mass moment of inertia in yaw 530 kgm
Wheel radius 0.43 m
TIP5-CT-2006-031312 Page 117 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
a.i-4 Vehicles suspension
The vehicles suspension elements are mainly divided in the primary and secondary suspension: the
primary suspension links the bogie to the axles while the secondary suspension links the bogie to the car
body.
Primary suspension Longitudinal stiffness 1.4 MN/m
Lateral stiffness 1.4 MN/m
Vertical stiffness 1.2 MN/m
Longitudinal damping 5.0 kNs/m
Lateral damping 5.0 kNs/m
Vertical damping 5.0 kNs/m
Secondary suspension Lateral stiffness 200 kN/m
Vertical stiffness 300 kN/m
Lateral damping Viscous damping (see figure below)
Vertical damping in compression = 2.3 kNs/m
in traction = 5.7 kNs/m
Yaw damping 5.0 kNs/m
TIP5-CT-2006-031312 Page 118 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Figure 5.2.3 Behaviour of the secondary lateral damper
Figure 5.2.4 Behaviour of the secondary vertical damper
TIP5-CT-2006-031312 Page 119 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii) Tramway vehicle
The tramway vehicle is the T3000 built by Bombardier for the STIB in Brussels. The figure below
represents the vehicle.
Figure 5.2.5 Drawings of the T3000 STIB Tramway
a.ii-1 Vehicle car body
Important parameters for the multi body of the tramway vehicles car body are listed below.
Mass (fully loaded) 47.9*10 kg
Mass moment of inertia in roll 15*10 kgm
Mass moment of inertia in pitch 390*10 kgm
Mass moment of inertia in yaw 390*10 kgm
Vertical location of centre of gravity above railhead 1.38 m
Longitudinal distance between bogie centres 9.369 m
a.ii-2 Vehicle bogies (without axles)
Important parameters for the multi body of the tramway vehicles bogies without axles are listed below.
Mass 3116 kg
Mass moment of inertia in roll 1722 kgm
Mass moment of inertia in pitch 1476 kgm
Mass moment of inertia in yaw 3067 kgm
Vertical location of centre of gravity above railhead 0.3 m
Longitudinal distance between wheel sets 1.85 m
TIP5-CT-2006-031312 Page 120 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
a.ii-3 Vehicle axles
Important parameters for the multi body of the light rail vehicles axles are listed below.
Mass 733.5 kg
Mass moment of inertia in roll 810 kgm
Mass moment of inertia in pitch 112 kgm
Mass moment of inertia in yaw 810 kgm
Wheel radius 0.29 m
a.ii-4 Vehicles suspension
The vehicles suspension elements are mainly divided in the primary and secondary suspension the
primary suspension links the bogie to the axles while the secondary suspension links the bogie to the car
body.
Primary suspension Longitudinal stiffness 1.84 MN/m
Lateral stiffness 1.84 MN/m
Vertical stiffness 1.58 MN/m
Longitudinal damping 5.0 kNs/m
Lateral damping 5.0 kNs/m
Vertical damping 5.0 kNs/m
Secondary suspension Lateral stiffness 200 kN/m
Vertical stiffness 300 kN/m
Lateral damping Viscous damping (see figure below)
Vertical damping in compression = 2.3 kNs/m
in traction = 5.7 kNs/m
Yaw damping 5.0 kNs/m
As no data was available on the damping of the suspension, the same values as for the light rail vehicle
were used.
TIP5-CT-2006-031312 Page 121 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Figure 5.2.6 Behaviour of the secondary lateral damper
Figure 5.2.7 Behaviour of the secondary vertical damper
TIP5-CT-2006-031312 Page 122 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(b) PRESENTATION OF THE TRACK MODELS
The track is modelled twice: once in Vi-Rail for the multi-body simulation and once in MSC. Nastran for
the calculation of stresses in the rail.
Track models vary in function of the support used ballasted track, continuously supported rail on
concrete slab or directly fastened rail on concrete slab.
Track models vary also in function of the type of rail used: NP4am, EB50T, 35G, Ri59N or Ri53N for the
tramway vehicle and UIC50 or UIC54 for the light rail vehicle.
Finally, the track models vary also in function of their layout: curved track (30 m radius curve for the
tramway vehicle and 100 m radius curve for the tramway vehicle) or straight track.
(i) Rail support type
The track model in Vi-Rail is a lumped parameters model as shown on the figure below.
Figure 5.2.8 Schematic view of the flexible track model in Vi-Rail
The parameters presented in this section are directly introduced in the model. Because the parts under
the base bushing (see dummy parts here above) are fixed, it is not possible to extract there displacement.
Therefore, displacements are measured above the sleeper and under the rail bushings. In the case of the
ballasted track, the base bushing corresponds to the ballast. In the case of direct fixation and continuously
supported track, there is no sleeper and no ballast. The base bushing stiffness is thus set to a very high
value. For uniformity among the simulation, the base bushing for the ballasted track was also set to a
very high value and instead, in the post-processing procedures, a filter representing the ballast is applied
to the vibration levels measured under the rail bushings. This filter is represented on the figure below.
TIP5-CT-2006-031312 Page 123 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Filter for ballast
-30
-25
-20
-15
-10
-5
0
5
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
freq (Hz)
d
B
[
]
Figure 5.2.9 Frequency spectrum of the filter used to model the ballast
b.i-1 Ballasted track
The ballasted track is rather standard the rail is fixed on a wooden sleeper through a rail pad and the
sleepers are placed on the ballast. The parameters of the different elements constituting the model are
listed below:
Rail Youngs modulus E 210 GPa
Poissons ratio 0.25
Density 7850
Rail pad Lateral stiffness 100 kN/mm
Lateral damping 2*10
4
Ns/m
Vertical stiffness 200 kN/mm
Vertical damping 2*10
4
Ns/m
Ballast Lateral stiffness 34 kN/mm
Lateral damping 2.5*10
5
Ns/m
Vertical stiffness 67 kN/mm
Vertical damping 1*10
6
Ns/m
Two distances between sleepers were used 60 cm and 75 cm.
TIP5-CT-2006-031312 Page 124 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.i-2 Continuously supported track
The continuously supported track is based on the CDM embedded rail system. Two systems were
modelled: the CDM Classic system and the CDM Comfort system. The rail has the same material
properties as for the ballasted track. The system is directly fastened in a concrete slab. The values of
stiffnesss listed below have been measured by Polimi during their procedure for fatigue tests on CDM
samples.
CDM Classic Lateral stiffness 50 kN/mm/lmrail
Lateral damping 2*10
4
Ns/m
Vertical stiffness 105 kN/mm/lmrail
Vertical damping 2*10
4
Ns/m
CDM Comfort Lateral stiffness 45 kN/mm/lmrail
Lateral damping 2*10
4
Ns/m
Vertical stiffness 50 kN/mm/lmrail
Vertical damping 2*10
4
Ns/m
b.i-3 Direct fixation track
The track with a rail directly fastened to a concrete slab has also been modelled with two different rail
pad stiffnesss.
Hard rail pad Lateral stiffness 60 kN/mm
Lateral damping 2*10
3
Ns/m
Vertical stiffness 185 kN/mm
Vertical damping 2*10
3
Ns/m
Soft fixation Lateral stiffness 40 kN/mm
Lateral damping 4*10
3
Ns/m
Vertical stiffness 50 kN/mm
Vertical damping 4*10
3
Ns/m
As for the ballasted track, two distances between fixations were used 60 cm and 75 cm.
TIP5-CT-2006-031312 Page 125 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii) Rail types
Seven types of rails were used in the simulations: five for the tramway vehicle and two for the light rail
vehicle.
b.ii-1 Rail types for the light rail vehicle
b.ii-1.1 UIC50
The figure below shows the parameters used to model the UIC50 rail as well as a drawing of the rail
section.
Figure 5.2.10 Drawing of the UIC50 rail section
Mass [kg/mrail] 50.46
IXX 1.532*10
-6
IYY 3.055*10
-6
IZZ 1.785*10
-5
Table 5.2.1 Parameters of the UIC50 rail
TIP5-CT-2006-031312 Page 126 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.ii-1.2 UIC54
Figure 5.2.11 Drawing of the UIC54 rail section
Mass [kg/mrail] 54.77
IXX 1.915*10
-6
IYY 4.208*10
-6
IZZ 2.334*10
-5
Table 5.2.2 Parameters of the UIC54 rail
TIP5-CT-2006-031312 Page 127 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.ii-2 Rail types for the tramway vehicle
b.ii-2.1 EB50T
Figure 5.2.12 Drawing of the EB50T rail section
Mass [kg/mrail] 50.1
IXX 1.358*10
-6
IYY 4.117*10
-6
IZZ 1.985*10
-5
Table 5.2.3 Parameters of the EB50T rail
TIP5-CT-2006-031312 Page 128 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.ii-2.2 NP4am
Figure 5.2.13 Drawing of the NP4am rail section
Mass [kg/mrail] 61.45
IXX 1.352*10
-6
IYY 1.042*10
-5
IZZ 3.466*10
-5
Table 5.2.4 Parameters of the NP4am rail
TIP5-CT-2006-031312 Page 129 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.ii-2.3 35G
Figure 5.2.14 Drawing of the NP4am rail section
Mass [kg/mrail] 61.45
IXX 1.26*10
-6
IYY 6.915*10
-6
IZZ 2.0715*10
-5
Table 5.2.5 Parameters of the 35G rail
TIP5-CT-2006-031312 Page 130 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.ii-2.4 Ri59N
Figure 5.2.15 Drawing of the 35G rail section
Mass [kg/mrail] 58.14
IXX 1.288*10
-6
IYY 8.819*10
-6
IZZ 3.281*10
-5
Table 5.2.6 Parameters of the Ri59N rail
TIP5-CT-2006-031312 Page 131 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
b.ii-2.5 Ri53N
Figure 5.2.16 Drawing of the Ri53N rail section
Mass [kg/mrail] 52.97
IXX 1.278*10
-6
IYY 7.449*10
-6
IZZ 1.344*10
-5
Table 5.2.7 Parameters of the Ri53N rail
TIP5-CT-2006-031312 Page 132 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(c) PRESENTATION OF THE TRACK LAYOUT
For each vehicle type (light rail and tramway), there are two tracks a straight track and a curved track.
Each track has a given length, curvature, rail inclination, gauge and rail irregularities.
Track length
The length of the track for both the light rail vehicle and the tramway vehicle is 500.0 m whether it is
straight or curved.
Track curvature radius
When the track s curved, two curvature radii have been chosen in function of the vehicle type:
for the light rail vehicle, the curvature radius is 100.0 m;
for the tramway vehicle, the curvature radius is 30.0 m.
Rail inclination
The inclination of the rails is the same for every track 0.025 rad.
Track Gauge
The track gauge is the same for every track 1.435 m.
Rail irregularities
Two types of rail irregularities have been modelled in Vi-Rail.
TIP5-CT-2006-031312 Page 133 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(i) Measured irregularities
A rail roughness that has been measured on a track in Kingston (Canada) has been used. The rail
roughness is displayed on the figure below.
Figure 5.2.17 Measured rail roughness spectrum
In Vi-Rail, the irregularities in displacement over distance on the rail can be directly inserted in the
model.
TIP5-CT-2006-031312 Page 134 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii) Step function irregularity
Another type of irregularities has been modelled in Vi-Rail: a step in the rail. The dimensions of this step
are as follows:
height of the step: 3 mm.
length of the step: 10 mm.
The figure below shows a representation of the modelled step.
Step irregularity
0
0.0005
0.001
0.0015
0.002
0.0025
0.003
0.0035
0 0.01 0.11 0.2
Length [m]
H
e
i
g
h
t
[
m
]
Figure 5.2.18 Representation of the step irregularity
TIP5-CT-2006-031312 Page 135 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
5.2.1.4. Results
Following the methodology, results are obtained at different stages of the chain. In the following sub-
sections, examples of results are presented. However, for size purposes, the major part of the results is
presented in the appendices.
(a) RESULTS OUT OF VI-RAIL (TIME DOMAIN)
After the calculation in Vi-Rail, track displacement and wheel/rail contact forces are plotted in function
of time.
(i) Displacements
On the figure below the displacement above the ballast and of the rails are displayed.
Figure 5.2.19 Example of displacements computed in Vi-Rail
The displacements under the track shown here above are post-processed in Matlab to obtain the
frequency spectrum of the velocity of vibration. Indeed, as rail irregularities are inserted in the model (in
the case of the figure above, measured irregularities); vibrations are generated by the vehicle passage.
TIP5-CT-2006-031312 Page 136 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii) Forces
The figure below shows the forces applied on the rail coming from the wheel/rail contact in function of
time.
Figure 5.2.20 Example of forces computed in Vi-Rail
Four forces are displayed the vertical and lateral forces on both the inner rail and the outer rail. On the
picture above, the forces presented come from the light rail vehicle running on the ballasted track with
EB50T rail on the curved track. The evolution of the forces over time clearly shows the vehicle entering
and getting out of the curve. One can also see the effect of the rail irregularities (here measured
irregularities).
For the vertical force, in the beginning, the force is the same on both rails and when the vehicle enters the
curve, the forces on the outer (inner) rail increases (decreases) to a certain constant level in the curve.
When the vehicle gets out of the curve, the forces go back to the same level as before the curve
corresponding to the axle load. The fact the vertical forces increase on the outer rail and decreases on the
inner rail can be explained by the fact that there is no cant deficiency imposed to the track. If there was,
one would expect the vertical forces to increase on the inner rail and decrease on the outer rail.
The force is calculated and stored in a spreadsheet (see table 4.2.3). This value of the force does not take
into account variations due to the rail irregularities because it is an average value over time. For the cases
where the vehicle runs in a curve the average of the force is calculated on the section of the time signal
where the vehicle is in the curve and not on the whole signal. The values of the forces are further
discussed below (see section 5.2.1.4.2.3).
TIP5-CT-2006-031312 Page 137 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(b) RESULTS OUT OF MATLAB
The displacement signals out of Vi-Rail are post-processed in Matlab to obtain frequency spectra of the
vibrations under the track.
(i) Vibration velocity spectra
The time signals of the displacements under the rail pads coming from Vi-Rail are transformed into
frequency spectra through a Fast Fourier Transform (FFT) algorithm and derived to obtain the velocity.
The spectra are then plotted in third-octave band as shown on the figure below. The resulting spectra
allow the comparison of vibration velocity spectra for the various combinations of parameters studied in
this report. The results are presented twice as two types of rail irregularities were used in the simulation
as explained in the above section: measured rail irregularities and a step function irregularity
(ii) Measured rail irregularities
In this section the vibration spectra obtained by simulation in Vi-Rail with measured rail irregularities are
discussed for some cases that cover the trends seen in all the results. The complete results are presented
in appendix A5.2.1 of the report.
Straight line on Ballast (UIC50)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
Figure 5.2.21 Vibration velocity spectra for the light rail vehicle in a straight line on ballast with UIC50 rail and
measured rail irregularities
On the figure above, the influence of the vehicle speed is clearly visible. The faster the vehicle, the highest
the vibrations are. However, this influence is much more visible at lower and higher frequencies: for the
frequencies around 32 Hz, the speed almost has no influence on the vibration levels.
TIP5-CT-2006-031312 Page 138 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Curve on ballast (UIC50)
-30
-20
-10
0
10
20
30
40
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Figure 5.2.22 Vibration velocity spectra for the light rail vehicle in a curve on ballast with UIC50 rail and
measured rail irregularities
The figure here above lead to another interesting remark: when the vehicle runs at 40 km/h and 60 km/h
peaks appear in the spectra at lower frequencies (8 Hz at 40 km/h and 13 Hz at 60 km/h). The lateral
movement of the bogies in the curve that occur when the vehicle runs faster can explain these peaks.
These movement are alternative movements that cause vibrations at a certain frequency and this
frequency increases when the vehicle runs faster because the bogies movement will occur more
frequently. At 20 km/h, apparently, the vehicle does not run fast enough for these movement to occur
but when it runs at 40 km/h, lateral movement of the bogie occur and they are even more frequent
(increased frequency from 8 to 13 Hz) and larger (higher peak at 13 Hz in the curve and 60 km/h).
TIP5-CT-2006-031312 Page 139 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Straight line at 20 km/h (UIC50)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Figure 5.2.23 Vibration velocity spectra for the light rail vehicle in a straight line with a UIC50 rail at 20 m/h for
various rail support types and measured rail irregularities
The figure here above allows comparing the performances of the different rail support types studied in
this report with respect to vibration mitigation. The ballasted track gives clearly the lowest levels but it is
important to remember that an attenuation filter was used to represent the ballast that had not been taken
into account in the flexible track in Vi-Rail. The chosen filter thus influences the spectrum for the
ballasted track. The direct fixation system with a hard railpad gives the highest vibration levels,
especially at higher frequencies. The CDM-Classic and the direct fixation with a soft railpad give very
similar results and the CDM-Comfort gives vibration levels that are slightly lower.
TIP5-CT-2006-031312 Page 140 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(iii) Step function rail irregularity
In this section, the vibration spectra obtained by simulation in Vi-Rail with a step function irregularity are
discussed for some cases that cover the trends seen in all the results. The complete results are presented
in appendix A5.2.2 of the report.
Straight line on CDM-Confort (UIC50)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
Figure 5.2.24 Vibration velocity spectra for the light rail vehicle in a straight line on a CDM-Comfort track with a
UIC50 rail and a step function irregularity
Unlike in the case of measured irregularities, the influence of the speed in the vibration spectrum due to a
step function irregularity is much more visible in all frequency bands. The vibration levels almost vary
with a factor of 20 * log10(V2/V1): between 20 and 40 km/h, the difference is approximately 6 dB while
between 20 and 80 km/h the difference is approximately 15 dB (not far from the 12 dB foreseen by the
above formula).
TIP5-CT-2006-031312 Page 141 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
Straight line at 60 km/h (UIC50)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Figure 5.2.25 Vibration velocity spectra for the light rail vehicle in a straight line on different types of track with a
UIC50 rail at 60 km/h and a step function irregularity
The figure above shows a comparison between the vibration spectra under the different types of track.
The same conclusion as for the results with a measured irregularity can be drawn to the exception that
the differences between the track systems are more visible.
TIP5-CT-2006-031312 Page 142 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(c) RESULTS OUT OF NASTRAN
The contact forces calculated in Vi-Rail are used as input forces in Nastran to calculate the stresses in the
rail web and in the rail foot.
The figure below shows the model of the EB50T rail on the ballasted track. The wheel/rail contact force
calculated in Vi-Rail is applied on the railhead.
Figure 5.2.26 Ballasted track Finite Element Model
The whole model is composed of 3D Hex elements. The rail pad (for ballast and discrete systems) or the
rubber materials (for the CDM system) are characterized by their Young moduli, which have been
numerically calibrated to obtain the same stiffness as those measured by Polimi and presented in section
2.2. The table below shows the stiffnesss Young moduli used in the FE models.
Type of track Support vertical stiffness
[kN/mm]
Support horizontal stiffness
[kN/mm]
Youngs modulus
[MPa]
Ballast track 200 100 28.4
Direct soft 50 40 7.1
Direct hard 185 60 26.27
CDM classic 63 30 1.47
CDM Comfort 50 27 7.1
Table 5.2.8 Parameters of the track Finite Element Models
The Von Mises stresses are calculated in every node of the mesh. The maximum stress in the rail web and
in the rail foot are saved in an excel file.
TIP5-CT-2006-031312 Page 143 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(i) Stresses in the rail web
Among the stresses in all the nodes of the rail, the maximum is taken among the nodes located between
two heights with reference to the bottom of the rail. The table below indicates the limits between which
the nodes are considered for calculation of the maximum Von Mises stress.
Type of rail Minimum node height
[mm]
Maximum node height
[mm]
EB50T 28 107
NP4am 20 117
35G 20 90
Ri59N 20 61
Ri53N 20 111
UIC50 28 103
UIC54 30 110
Table 5.2.9 Location of the nodes where the maximum stress is computed in the rail web
The figure below shows an example of the Von Mises stresses in the EB50T rail on the ballasted track
under the load of the tramway vehicle in a curve at 20 km/h.
Figure 5.2.27
TIP5-CT-2006-031312 Page 144 of 289
URBAN TRACK Issued: 15/11/2010
D0602_M48_UITP_SP5.doc
(ii) Stresses in the rail foot
Unlike for the stresses in the rail web, the maximum stress in the rail foot is calculated among the nodes
located on the bottom of the rail.
The figure below shows an example of the Von Mises stresses in the EB50T rail on the ballasted track
under the load of the tramway vehicle in a curve at 20 km/h.
Figure 5.2.28
(iii) Conclusions on the stresses in the rail
Finally, the two tables below summarize the forces applied on the rails and the stresses in the rail for each
simulation case.
TIP5-CT-2006-031312 Page 145 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Outer rail Inner rail 60 cm 75 cm Continuous
Force Force Max stress Max stress Max stress
Vehicle Rail
profile
Curve
radius
Fixation
system
V
e
h
i
c
l
e
v
e
l
o
c
i
t
y
Vertical Lateral Vertical Lateral web foot web foot web foot
STIB_T3000 EB50T 30m Ballast 20 63539.16 -38067.78 40454.20 -13902.56 149.86 71.35 150.73 81.78 0.00 0.00
STIB_T3000 EB50T 30m CDM_Classic 20 68222.92 -35599.44 35808.93 -12281.68 0.00 0.00 0.00 0.00 72.06 30.02
STIB_T3000 EB50T 30m CDM_Confort 20 67260.91 -36126.16 36774.44 -12650.87 0.00 0.00 0.00 0.00 63.91 29.76
STIB_T3000 EB50T 30m Direct_Hard 20 69383.37 -34866.54 34649.58 -11775.02 140.45 67.66 140.79 76.57 0.00 0.00
STIB_T3000 EB50T 30m Direct_Soft 20 67866.54 -35707.48 36165.38 -12366.55 140.83 70.38 147.06 79.59 0.00 0.00
STIB_T3000 EB50T Infinite Ballast 20 51977.90 128.55 51977.90 128.55 34.68 24.58 34.38 26.76 0.00 0.00
STIB_T3000 EB50T Infinite Ballast 40 51977.90 128.55 51977.90 128.55 34.68 24.58 34.38 26.76 0.00 0.00
STIB_T3000 EB50T Infinite Ballast 60 51977.90 128.55 51977.90 128.55 34.68 24.58 34.38 26.76 0.00 0.00
STIB_T3000 EB50T Infinite CDM_Classic 20 51977.90 127.36 51977.90 127.36 0.00 0.00 0.00 0.00 35.35 16.86
STIB_T3000 EB50T Infinite CDM_Classic 40 51977.90 127.36 51977.90 127.36 0.00 0.00 0.00 0.00 35.35 16.86
STIB_T3000 EB50T Infinite CDM_Classic 60 51977.90 127.36 51977.90 127.36 0.00 0.00 0.00 0.00 35.35 16.86
STIB_T3000 EB50T Infinite CDM_Confort 20 51978.40 127.13 51977.40 127.11 0.00 0.00 0.00 0.00 35.25 17.20
STIB_T3000 EB50T Infinite CDM_Confort 40 51978.40 127.13 51977.40 127.11 0.00 0.00 0.00 0.00 35.25 17.20
STIB_T3000 EB50T Infinite CDM_Confort 60 51978.40 127.13 51977.40 127.11 0.00 0.00 0.00 0.00 35.25 17.20
STIB_T3000 EB50T Infinite Direct_Hard 20 51977.90 128.11 51977.90 128.10 34.72 24.09 34.37 26.74 0.00 0.00
STIB_T3000 EB50T Infinite Direct_Hard 40 51977.90 128.11 51977.90 128.10 34.72 24.09 34.37 26.74 0.00 0.00
STIB_T3000 EB50T Infinite Direct_Hard 60 51977.90 128.11 51977.90 128.10 34.72 24.09 34.37 26.74 0.00 0.00
STIB_T3000 EB50T Infinite Direct_Soft 20 51977.90 127.91 51977.90 127.91 34.69 24.35 34.37 27.17 0.00 0.00
STIB_T3000 EB50T Infinite Direct_Soft 40 51977.90 127.91 51977.90 127.91 34.69 24.35 34.37 27.17 0.00 0.00
STIB_T3000 EB50T Infinite Direct_Soft 60 51977.90 127.91 51977.90 127.91 34.69 24.35 34.37 27.17 0.00 0.00
STIB_T3000 NP4am 30m Ballast 20 64796.26 -36242.01 39171.51 -12762.06 208.98 129.98 224.63 137.40 0.00 0.00
STIB_T3000 NP4am 30m CDM_Classic 20 69657.09 -33308.64 34361.55 -10867.35 0.00 0.00 0.00 0.00 73.57 34.67
STIB_T3000 NP4am 30m CDM_Confort 20 68734.03 -33870.26 35287.63 -11257.28 0.00 0.00 0.00 0.00 74.05 38.00
STIB_T3000 NP4am 30m Direct_Hard 20 70640.61 -32590.57 33379.59 -10375.88 186.14 126.18 200.01 131.38 0.00 0.00
STIB_T3000 NP4am 30m Direct_Soft 20 69235.29 -33444.68 34782.49 -10971.39 192.16 134.01 207.72 139.71 0.00 0.00
TIP5-CT-2006-031312 Page 146 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Outer rail Inner rail 60 cm 75 cm Continuous
Force Force Max stress Max stress Max stress
Vehicle Rail
profile
Curve
radius
Fixation
system
V
e
h
i
c
l
e
v
e
l
o
c
i
t
y
Vertical Lateral Vertical Lateral web foot web foot web foot
STIB_T3000 NP4am Infinite Ballast 20 51977.90 -403.53 51977.90 -403.53 35.59 41.84 36.27 43.65 0.00 0.00
STIB_T3000 NP4am Infinite Ballast 40 51977.90 -403.53 51977.90 -403.53 35.59 41.84 36.27 43.65 0.00 0.00
STIB_T3000 NP4am Infinite Ballast 60 51977.90 -403.53 51977.90 -403.53 35.59 41.84 36.27 43.65 0.00 0.00
STIB_T3000 NP4am Infinite CDM_Classic 20 51977.90 -397.15 51977.90 -397.15 0.00 0.00 0.00 0.00 37.36 20.95
STIB_T3000 NP4am Infinite CDM_Classic 40 51977.90 -397.15 51977.90 -397.15 0.00 0.00 0.00 0.00 37.36 20.95
STIB_T3000 NP4am Infinite CDM_Classic 60 51977.90 -397.15 51977.90 -397.15 0.00 0.00 0.00 0.00 37.36 20.95
STIB_T3000 NP4am Infinite CDM_Confort 20 51978.50 -396.32 51977.30 -396.38 0.00 0.00 0.00 0.00 39.21 24.78
STIB_T3000 NP4am Infinite CDM_Confort 40 51978.50 -396.32 51977.30 -396.38 0.00 0.00 0.00 0.00 39.21 24.78
STIB_T3000 NP4am Infinite CDM_Confort 60 51978.50 -396.32 51977.30 -396.38 0.00 0.00 0.00 0.00 39.21 24.78
STIB_T3000 NP4am Infinite Direct_Hard 20 51977.90 -402.05 51977.90 -402.06 35.60 41.77 36.28 43.56 0.00 0.00
STIB_T3000 NP4am Infinite Direct_Hard 40 51977.90 -402.05 51977.90 -402.06 35.60 41.77 36.28 43.56 0.00 0.00
STIB_T3000 NP4am Infinite Direct_Hard 60 51977.90 -402.05 51977.90 -402.06 35.60 41.77 36.28 43.56 0.00 0.00
STIB_T3000 NP4am Infinite Direct_Soft 20 51977.90 -398.97 51977.90 -398.97 35.62 45.72 36.29 47.54 0.00 0.00
STIB_T3000 NP4am Infinite Direct_Soft 40 51977.90 -398.97 51977.90 -398.97 35.62 45.72 36.29 47.54 0.00 0.00
STIB_T3000 NP4am Infinite Direct_Soft 60 51977.90 -398.97 51977.90 -398.97 35.62 45.72 36.29 47.54 0.00 0.00
STIB_T3000 Ri59N 30m Ballast 20 64796.29 -36242.08 39171.47 -12762.06 277.50 132.94 295.15 142.75 0.00 0.00
STIB_T3000 Ri59N 30m CDM_Classic 20 69657.08 -33308.61 34361.57 -10867.34 0.00 0.00 0.00 0.00 83.11 39.51
STIB_T3000 Ri59N 30m CDM_Confort 20 68735.18 -33872.62 35286.49 -11257.46 0.00 0.00 0.00 0.00 35.69 26.22
STIB_T3000 Ri59N 30m Direct_Hard 20 70640.64 -32590.62 33379.55 -10375.88 253.39 127.05 269.16 134.71 0.00 0.00
STIB_T3000 Ri59N 30m Direct_Soft 20 69235.01 -33444.15 34782.77 -10971.35 261.27 136.59 278.20 144.18 0.00 0.00
STIB_T3000 Ri59N Infinite Ballast 20 51977.90 -403.53 51977.90 -403.53 30.71 41.82 31.59 43.80 0.00 0.00
STIB_T3000 Ri59N Infinite Ballast 40 51977.90 -403.53 51977.90 -403.53 30.71 41.82 31.59 43.80 0.00 0.00
STIB_T3000 Ri59N Infinite Ballast 60 51977.90 -403.53 51977.90 -403.53 30.71 41.82 31.59 43.80 0.00 0.00
STIB_T3000 Ri59N Infinite CDM_Classic 20 51977.90 -397.15 51977.90 -397.15 0.00 0.00 0.00 0.00 32.04 22.63
STIB_T3000 Ri59N Infinite CDM_Classic 40 51977.90 -397.15 51977.90 -397.15 0.00 0.00 0.00 0.00 32.04 22.63
TIP5-CT-2006-031312 Page 147 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Outer rail Inner rail 60 cm 75 cm Continuous
Force Force Max stress Max stress Max stress
Vehicle Rail
profile
Curve
radius
Fixation
system
V
e
h
i
c
l
e
v
e
l
o
c
i
t
y
Vertical Lateral Vertical Lateral web foot web foot web foot
STIB_T3000 Ri59N Infinite CDM_Classic 60 51977.90 -397.15 51977.90 -397.15 0.00 0.00 0.00 0.00 32.04 22.63
STIB_T3000 Ri59N Infinite CDM_Confort 20 51978.50 -396.32 51977.40 -396.38 0.00 0.00 0.00 0.00 35.69 26.22
STIB_T3000 Ri59N Infinite CDM_Confort 40 51978.50 -396.32 51977.40 -396.38 0.00 0.00 0.00 0.00 35.69 26.22
STIB_T3000 Ri59N Infinite CDM_Confort 60 51978.50 -396.32 51977.40 -396.38 0.00 0.00 0.00 0.00 35.69 26.22
STIB_T3000 Ri59N Infinite Direct_Hard 20 51977.90 -402.05 51977.90 -402.06 30.72 41.74 31.60 43.70 0.00 0.00
STIB_T3000 Ri59N Infinite Direct_Hard 40 51977.90 -402.05 51977.90 -402.06 30.72 41.74 31.60 43.70 0.00 0.00
STIB_T3000 Ri59N Infinite Direct_Hard 60 51977.90 -402.05 51977.90 -402.06 30.72 41.74 31.60 43.70 0.00 0.00
STIB_T3000 Ri59N Infinite Direct_Soft 20 51977.90 -398.97 51977.90 -398.97 30.76 45.29 31.62 47.29 0.00 0.00
STIB_T3000 Ri59N Infinite Direct_Soft 40 51977.90 -398.97 51977.90 -398.97 30.76 45.29 31.62 47.29 0.00 0.00
STIB_T3000 Ri59N Infinite Direct_Soft 60 51977.90 -398.97 51977.90 -398.97 30.76 45.29 31.62 47.29 0.00 0.00
STIB_T3000 35G 30m Ballast 20 64796.27 -36242.02 39171.51 -12762.06 195.61 91.41 204.48 99.07 0.00 0.00
STIB_T3000 35G 30m CDM_Classic 20 69657.09 -33308.64 34361.55 -10867.35 0.00 0.00 0.00 0.00 81.93 29.90
STIB_T3000 35G 30m CDM_Confort 20 68733.21 -33868.61 35288.46 -11257.15 0.00 0.00 0.00 0.00 79.16 29.82
STIB_T3000 35G 30m Direct_Hard 20 70640.61 -32590.57 33379.59 -10375.88 180.43 88.07 187.29 94.97 0.00 0.00
STIB_T3000 35G 30m Direct_Soft 20 69235.29 -33444.68 34782.49 -10971.39 183.94 90.62 192.15 97.97 0.00 0.00
STIB_T3000 35G Infinite Ballast 20 51977.90 -403.53 51977.90 -403.53 38.78 31.69 38.78 34.21 0.00 0.00
STIB_T3000 35G Infinite Ballast 40 51977.90 -403.53 51977.90 -403.53 38.78 31.69 38.78 34.21 0.00 0.00
STIB_T3000 35G Infinite Ballast 60 51977.90 -403.53 51977.90 -403.53 38.78 31.69 38.78 34.21 0.00 0.00
STIB_T3000 35G Infinite CDM_Classic 20 51977.90 -397.15 51977.90 -397.15 0.00 0.00 0.00 0.00 38.03 18.82
STIB_T3000 35G Infinite CDM_Classic 40 51977.90 -397.15 51977.90 -397.15 0.00 0.00 0.00 0.00 38.03 18.82
STIB_T3000 35G Infinite CDM_Classic 60 51977.90 -397.15 51977.90 -397.15 0.00 0.00 0.00 0.00 38.03 18.82
STIB_T3000 35G Infinite CDM_Confort 20 51978.50 -396.32 51977.30 -396.38 0.00 0.00 0.00 0.00 38.49 19.41
STIB_T3000 35G Infinite CDM_Confort 40 51978.50 -396.32 51977.30 -396.38 0.00 0.00 0.00 0.00 38.49 19.41
STIB_T3000 35G Infinite CDM_Confort 60 51978.50 -396.32 51977.30 -396.38 0.00 0.00 0.00 0.00 38.49 19.41
STIB_T3000 35G Infinite Direct_Hard 20 51977.90 -402.05 51977.90 -402.06 38.78 31.64 38.77 34.16 0.00 0.00
TIP5-CT-2006-031312 Page 148 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Outer rail Inner rail 60 cm 75 cm Continuous
Force Force Max stress Max stress Max stress
Vehicle Rail
profile
Curve
radius
Fixation
system
V
e
h
i
c
l
e
v
e
l
o
c
i
t
y
Vertical Lateral Vertical Lateral web foot web foot web foot
STIB_T3000 35G Infinite Direct_Hard 40 51977.90 -402.05 51977.90 -402.06 38.78 31.64 38.77 34.16 0.00 0.00
STIB_T3000 35G Infinite Direct_Hard 60 51977.90 -402.05 51977.90 -402.06 38.78 31.64 38.77 34.16 0.00 0.00
STIB_T3000 35G Infinite Direct_Soft 20 51977.90 -398.97 51977.90 -398.97 38.78 32.78 38.77 35.33 0.00 0.00
STIB_T3000 35G Infinite Direct_Soft 40 51977.90 -398.97 51977.90 -398.97 38.78 32.78 38.77 35.33 0.00 0.00
STIB_T3000 35G Infinite Direct_Soft 60 51977.90 -398.97 51977.90 -398.97 38.78 32.78 38.77 35.33 0.00 0.00
STIB_T3000 Ri53N 30m Ballast 20 64796.34 -36242.18 39171.43 -12762.06 158.99 102.95 163.63 109.19 0.00 0.00
STIB_T3000 Ri53N 30m CDM_Classic 20 69657.97 -33310.18 34360.71 -10867.43 0.00 0.00 0.00 0.00 69.93 39.29
STIB_T3000 Ri53N 30m CDM_Confort 20 68735.67 -33873.67 35286.00 -11257.53 0.00 0.00 0.00 0.00 70.57 39.96
STIB_T3000 Ri53N 30m Direct_Hard 20 70640.69 -32590.69 33379.50 -10375.88 145.90 100.15 149.59 105.82 0.00 0.00
STIB_T3000 Ri53N 30m Direct_Soft 20 69235.01 -33444.14 34782.77 -10971.35 148.99 104.46 153.49 110.36 0.00 0.00
STIB_T3000 Ri53N Infinite Ballast 20 51977.90 -403.53 51977.90 -403.53 33.31 42.88 33.60 47.15 0.00 0.00
STIB_T3000 Ri53N Infinite Ballast 40 51977.90 -403.53 51977.90 -403.53 33.31 42.88 33.60 47.15 0.00 0.00
STIB_T3000 Ri53N Infinite Ballast 60 51977.90 -403.53 51977.90 -403.53 33.31 42.88 33.60 47.15 0.00 0.00
STIB_T3000 Ri53N Infinite CDM_Classic 20 51977.90 -397.15 51977.90 -397.15 0.00 0.00 0.00 0.00 33.59 23.96
STIB_T3000 Ri53N Infinite CDM_Classic 40 51977.90 -397.15 51977.90 -397.15 0.00 0.00 0.00 0.00 33.59 23.96
STIB_T3000 Ri53N Infinite CDM_Classic 60 51977.90 -397.15 51977.90 -397.15 0.00 0.00 0.00 0.00 33.59 23.96
STIB_T3000 Ri53N Infinite CDM_Confort 20 51978.50 -396.32 51977.40 -396.38 0.00 0.00 0.00 0.00 36.49 25.23
STIB_T3000 Ri53N Infinite CDM_Confort 40 51978.50 -396.32 51977.40 -396.38 0.00 0.00 0.00 0.00 36.49 25.23
STIB_T3000 Ri53N Infinite CDM_Confort 60 51978.45 -396.32 51977.40 -396.38 0.00 0.00 0.00 0.00 36.49 25.23
STIB_T3000 Ri53N Infinite Direct_Hard 20 51977.90 -402.05 51977.90 -402.06 33.31 42.80 33.61 47.09 0.00 0.00
STIB_T3000 Ri53N Infinite Direct_Hard 40 51977.90 -402.05 51977.90 -402.06 33.31 42.80 33.61 47.09 0.00 0.00
STIB_T3000 Ri53N Infinite Direct_Hard 60 51977.90 -402.05 51977.90 -402.06 33.31 42.80 33.61 47.09 0.00 0.00
STIB_T3000 Ri53N Infinite Direct_Soft 20 51977.90 -398.97 51977.90 -398.97 33.30 44.54 33.61 48.88 0.00 0.00
STIB_T3000 Ri53N Infinite Direct_Soft 40 51977.90 -398.97 51977.90 -398.97 33.30 44.54 33.61 48.88 0.00 0.00
STIB_T3000 Ri53N Infinite Direct_Soft 60 51977.90 -398.97 51977.90 -398.97 33.30 44.54 33.61 48.88 0.00 0.00
TIP5-CT-2006-031312 Page 149 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Outer rail Inner rail 60 cm 75 cm Continuous
Force Force Max stress Max stress Max stress
Vehicle Rail
profile
Curve
radius
Fixation
system
V
e
h
i
c
l
e
v
e
l
o
c
i
t
y
Vertical Lateral Vertical Lateral web foot web foot web foot
Light_Rail UIC50 100m Ballast 20 51438.57 -32491.46 29715.73 -11916.84 153.52 70.29 160.09 79.99 0.00 0.00
Light_Rail UIC50 100m Ballast 40 61068.07 -36246.20 20094.32 -8023.59 172.25 79.19 179.38 90.20 0.00 0.00
Light_Rail UIC50 100m Ballast 60 75428.72 -39053.77 5742.40 -2290.54 188.20 87.39 195.53 99.77 0.00 0.00
Light_Rail UIC50 100m CDM_Classic 20 53594.19 -31201.63 27590.58 -11104.73 0.00 0.00 0.00 0.00 66.30 32.23
Light_Rail UIC50 100m CDM_Classic 40 63566.83 -36103.29 17650.61 -7058.08 0.00 0.00 0.00 0.00 75.29 37.59
Light_Rail UIC50 100m CDM_Classic 60 77156.75 -38812.43 4133.77 -1649.37 0.00 0.00 0.00 0.00 83.19 42.86
Light_Rail UIC50 100m CDM_Confort 20 53251.32 -31398.82 27934.64 -11248.37 0.00 0.00 0.00 0.00 57.48 32.13
Light_Rail UIC50 100m CDM_Confort 40 63124.60 -36086.71 18098.08 -7243.97 0.00 0.00 0.00 0.00 65.38 37.35
Light_Rail UIC50 100m CDM_Confort 60 76843.11 -38786.84 4460.16 -1783.21 0.00 0.00 0.00 0.00 72.60 42.63
Light_Rail UIC50 100m Direct_Hard 20 53950.86 -31005.24 27235.13 -10945.97 153.51 70.11 160.35 79.68 0.00 0.00
Light_Rail UIC50 100m Direct_Hard 40 64103.81 -36177.18 17114.65 -6832.18 172.76 79.54 180.03 90.32 0.00 0.00
Light_Rail UIC50 100m Direct_Hard 60 77578.43 -38922.85 3711.90 -1474.99 188.20 87.49 195.58 99.40 0.00 0.00
Light_Rail UIC50 100m Direct_Soft 20 53455.53 -31284.20 27728.13 -11152.73 153.52 70.24 161.12 79.79 0.00 0.00
Light_Rail UIC50 100m Direct_Soft 40 63446.61 -36146.33 17772.34 -7104.81 172.47 79.44 180.63 90.29 0.00 0.00
Light_Rail UIC50 100m Direct_Soft 60 77080.00 -38852.39 4220.49 -1682.91 187.79 87.29 196.14 99.26 0.00 0.00
Light_Rail UIC50 Infinite Ballast 20 40547.79 76.33 40547.78 76.33 25.60 17.94 25.26 20.50 0.00 0.00
Light_Rail UIC50 Infinite Ballast 40 40547.81 76.33 40547.83 76.33 25.60 17.94 25.26 20.50 0.00 0.00
Light_Rail UIC50 Infinite Ballast 60 40547.84 76.33 40547.80 76.33 25.60 17.94 25.26 20.50 0.00 0.00
Light_Rail UIC50 Infinite Ballast 80 40547.92 76.32 40547.83 76.33 25.60 17.94 25.26 20.50 0.00 0.00
Light_Rail UIC50 Infinite Ballast 100 40547.82 76.33 40547.82 76.33 25.60 17.94 25.26 20.50 0.00 0.00
Light_Rail UIC50 Infinite CDM_Classic 20 40547.79 75.70 40547.80 75.70 0.00 0.00 0.00 0.00 26.36 13.87
Light_Rail UIC50 Infinite CDM_Classic 40 40547.80 75.70 40547.80 75.70 0.00 0.00 0.00 0.00 26.36 13.87
Light_Rail UIC50 Infinite CDM_Classic 60 40547.81 75.70 40548.01 75.70 0.00 0.00 0.00 0.00 26.36 13.87
Light_Rail UIC50 Infinite CDM_Classic 80 40547.79 75.70 40547.78 75.70 0.00 0.00 0.00 0.00 26.36 13.87
Light_Rail UIC50 Infinite CDM_Classic 100 40547.79 75.70 40547.80 75.70 0.00 0.00 0.00 0.00 26.36 13.87
TIP5-CT-2006-031312 Page 150 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Outer rail Inner rail 60 cm 75 cm Continuous
Force Force Max stress Max stress Max stress
Vehicle Rail
profile
Curve
radius
Fixation
system
V
e
h
i
c
l
e
v
e
l
o
c
i
t
y
Vertical Lateral Vertical Lateral web foot web foot web foot
Light_Rail UIC50 Infinite CDM_Confort 20 40548.20 75.58 40547.40 75.56 0.00 0.00 0.00 0.00 26.27 14.02
Light_Rail UIC50 Infinite CDM_Confort 40 40548.20 75.58 40547.40 75.56 0.00 0.00 0.00 0.00 26.27 14.02
Light_Rail UIC50 Infinite CDM_Confort 60 40548.20 75.58 40547.40 75.56 0.00 0.00 0.00 0.00 26.27 14.02
Light_Rail UIC50 Infinite CDM_Confort 80 40548.20 75.58 40547.40 75.56 0.00 0.00 0.00 0.00 26.27 14.02
Light_Rail UIC50 Infinite CDM_Confort 100 40548.20 75.58 40547.40 75.56 0.00 0.00 0.00 0.00 26.27 14.02
Light_Rail UIC50 Infinite Direct_Hard 20 40547.80 76.08 40547.80 76.08 25.60 17.87 25.25 20.50 0.00 0.00
Light_Rail UIC50 Infinite Direct_Hard 40 40547.83 76.08 40547.81 76.08 25.60 17.87 25.25 20.50 0.00 0.00
Light_Rail UIC50 Infinite Direct_Hard 60 40547.80 76.08 40547.80 76.08 25.60 17.87 25.25 20.50 0.00 0.00
Light_Rail UIC50 Infinite Direct_Hard 80 40548.18 76.08 40547.79 76.08 25.60 17.87 25.25 20.50 0.00 0.00
Light_Rail UIC50 Infinite Direct_Hard 100 40547.82 76.08 40547.82 76.08 25.60 17.87 25.25 20.50 0.00 0.00
Light_Rail UIC50 Infinite Direct_Soft 20 40547.79 76.01 40547.79 76.01 25.59 17.94 25.24 20.76 0.00 0.00
Light_Rail UIC50 Infinite Direct_Soft 40 40547.81 76.01 40547.80 76.01 25.59 17.94 25.24 20.76 0.00 0.00
Light_Rail UIC50 Infinite Direct_Soft 60 40547.82 76.01 40547.81 76.01 25.59 17.94 25.24 20.76 0.00 0.00
Light_Rail UIC50 Infinite Direct_Soft 80 40547.81 76.01 40547.81 76.01 25.59 17.94 25.24 20.76 0.00 0.00
Light_Rail UIC50 Infinite Direct_Soft 100 40547.81 76.01 40547.81 76.01 25.59 17.94 25.24 20.76 0.00 0.00
Light_Rail UIC54 100m Ballast 20 51439.54 -32492.90 29715.31 -11916.64 102.06 57.75 106.76 66.50 0.00 0.00
Light_Rail UIC54 100m Ballast 40 61066.00 -36245.48 20093.32 -8023.19 114.58 64.71 119.68 74.50 0.00 0.00
Light_Rail UIC54 100m Ballast 60 75428.50 -39052.36 5733.72 -2287.08 125.40 71.51 130.58 81.74 0.00 0.00
Light_Rail UIC54 100m CDM_Classic 20 53594.38 -31201.69 27590.76 -11104.81 0.00 0.00 0.00 0.00 55.02 26.88
Light_Rail UIC54 100m CDM_Classic 40 63566.76 -36103.30 17650.85 -7058.11 0.00 0.00 0.00 0.00 64.06 31.44
Light_Rail UIC54 100m CDM_Classic 60 77167.14 -38815.87 4131.42 -1648.43 0.00 0.00 0.00 0.00 73.41 35.74
Light_Rail UIC54 100m CDM_Confort 20 53251.82 -31399.03 27934.57 -11248.35 0.00 0.00 0.00 0.00 51.82 26.82
Light_Rail UIC54 100m CDM_Confort 40 63125.51 -36086.97 18097.73 -7243.76 0.00 0.00 0.00 0.00 60.88 31.25
Light_Rail UIC54 100m CDM_Confort 60 76851.03 -38790.43 4461.40 -1783.73 0.00 0.00 0.00 0.00 71.65 35.56
Light_Rail UIC54 100m Direct_Hard 20 53950.82 -31005.24 27235.10 -10945.95 98.35 55.45 102.85 63.63 0.00 0.00
TIP5-CT-2006-031312 Page 151 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Outer rail Inner rail 60 cm 75 cm Continuous
Force Force Max stress Max stress Max stress
Vehicle Rail
profile
Curve
radius
Fixation
system
V
e
h
i
c
l
e
v
e
l
o
c
i
t
y
Vertical Lateral Vertical Lateral web foot web foot web foot
Light_Rail UIC54 100m Direct_Hard 40 64104.33 -36177.25 17114.07 -6831.87 114.99 64.92 120.20 74.33 0.00 0.00
Light_Rail UIC54 100m Direct_Hard 60 77578.73 -38922.65 3711.45 -1474.83 125.45 71.48 130.74 81.38 0.00 0.00
Light_Rail UIC54 100m Direct_Soft 20 53455.75 -31284.31 27728.36 -11152.82 99.05 55.92 104.06 64.19 0.00 0.00
Light_Rail UIC54 100m Direct_Soft 40 63446.49 -36146.29 17772.59 -7104.84 114.79 64.92 120.51 74.30 0.00 0.00
Light_Rail UIC54 100m Direct_Soft 60 77074.87 -38851.54 4219.45 -1682.51 125.18 71.43 131.01 81.39 0.00 0.00
Light_Rail UIC54 Infinite Ballast 20 40547.84 76.33 40547.82 76.33 28.15 15.53 27.87 17.66 0.00 0.00
Light_Rail UIC54 Infinite Ballast 40 40547.80 76.33 40547.80 76.33 28.15 15.53 27.87 17.66 0.00 0.00
Light_Rail UIC54 Infinite Ballast 60 40547.83 76.33 40547.79 76.33 28.15 15.53 27.87 17.66 0.00 0.00
Light_Rail UIC54 Infinite Ballast 80 40547.80 76.33 40547.80 76.33 28.15 15.53 27.87 17.66 0.00 0.00
Light_Rail UIC54 Infinite Ballast 100 40547.78 76.33 40547.80 76.33 28.15 15.53 27.87 17.66 0.00 0.00
Light_Rail UIC54 Infinite CDM_Classic 20 40547.81 75.70 40547.80 75.70 0.00 0.00 0.00 0.00 28.92 11.29
Light_Rail UIC54 Infinite CDM_Classic 40 40547.82 75.70 40547.83 75.70 0.00 0.00 0.00 0.00 28.92 11.29
Light_Rail UIC54 Infinite CDM_Classic 60 40547.78 75.70 40547.79 75.70 0.00 0.00 0.00 0.00 28.92 11.29
Light_Rail UIC54 Infinite CDM_Classic 80 40547.79 75.70 40547.80 75.70 0.00 0.00 0.00 0.00 28.92 11.29
Light_Rail UIC54 Infinite CDM_Classic 100 40547.80 75.70 40547.80 75.70 0.00 0.00 0.00 0.00 28.92 11.29
Light_Rail UIC54 Infinite CDM_Confort 20 40548.20 75.58 40547.40 75.56 0.00 0.00 0.00 0.00 28.85 11.47
Light_Rail UIC54 Infinite CDM_Confort 40 40548.20 75.58 40547.40 75.56 0.00 0.00 0.00 0.00 28.85 11.47
Light_Rail UIC54 Infinite CDM_Confort 60 40548.20 75.58 40547.40 75.56 0.00 0.00 0.00 0.00 28.85 11.47
Light_Rail UIC54 Infinite CDM_Confort 80 40548.20 75.58 40547.40 75.56 0.00 0.00 0.00 0.00 28.85 11.47
Light_Rail UIC54 Infinite CDM_Confort 100 40548.20 75.58 40547.40 75.56 0.00 0.00 0.00 0.00 28.85 11.47
Light_Rail UIC54 Infinite Direct_Hard 20 40547.79 76.08 40547.57 76.08 28.15 15.32 27.86 17.65 0.00 0.00
Light_Rail UIC54 Infinite Direct_Hard 40 40547.76 76.08 40547.81 76.08 28.15 15.32 27.86 17.65 0.00 0.00
Light_Rail UIC54 Infinite Direct_Hard 60 40547.82 76.08 40547.83 76.08 28.15 15.32 27.86 17.65 0.00 0.00
Light_Rail UIC54 Infinite Direct_Hard 80 40547.79 76.08 40547.79 76.08 28.15 15.32 27.86 17.65 0.00 0.00
Light_Rail UIC54 Infinite Direct_Hard 100 40547.78 76.09 40547.80 76.08 28.15 15.32 27.86 17.65 0.00 0.00
TIP5-CT-2006-031312 Page 152 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Outer rail Inner rail 60 cm 75 cm Continuous
Force Force Max stress Max stress Max stress
Vehicle Rail
profile
Curve
radius
Fixation
system
V
e
h
i
c
l
e
v
e
l
o
c
i
t
y
Vertical Lateral Vertical Lateral web foot web foot web foot
Light_Rail UIC54 Infinite Direct_Soft 20 40547.78 76.01 40547.81 76.01 28.15 15.52 27.85 17.87 0.00 0.00
Light_Rail UIC54 Infinite Direct_Soft 40 40547.82 76.01 40547.84 76.01 28.15 15.52 27.85 17.87 0.00 0.00
Light_Rail UIC54 Infinite Direct_Soft 60 40547.80 76.01 40547.80 76.01 28.15 15.52 27.85 17.87 0.00 0.00
Light_Rail UIC54 Infinite Direct_Soft 80 40547.85 76.01 40547.79 76.00 28.15 15.52 27.85 17.87 0.00 0.00
Light_Rail UIC54 Infinite Direct_Soft 100 40547.93 76.01 40547.87 76.01 28.15 15.52 27.85 17.87 0.00 0.00
Table 5.2.10 Results for all simulation cases (forces on the rail and stresses in the rail)
TIP5-CT-2006-031312 Page 153 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
Looking at the results presented in the tables above, sever conclusions can be drawn.
First, the wheel load applied on the rail is higher for the tramway than for the light rail. This is logical
because the tramway is heavier. Another important point is that the vertical load increases on the outer
rail in curve and decreases on the inner rail. This is also straightforward as there is no canting on the
track.
In straight line, the load almost does not vary with the speed, the support type or the rail type. It is
important to remember here that the load is the mean load so the influence of the irregularities on the rail
is not taken into account. Therefore, the load does not vary with the speed.
The stresses presented in the table are computed in Nastran as explained in the section presenting the
methodology. The stresses are classified in function of the distance between the fixations 60 cm, 75 cm
and continuously supported rail. Of course, for the ballasted track and the direct fixation track do not
have results for the continuously supported rail, as these are tracks where the rail is discretely supported.
On the contrary, the CDM_Classic and CDM_Comfort tracks only have results for a continuously
supported rail.
In curve, the stresses are a little higher in the rail on the ballasted track than in the direct fixation tracks.
The stresses in the rail on the hard direct fixation are a little lower than on the soft direct fixation. Another
interesting fact is that the stresses are higher when the distance between fixations is higher. This is logical
since the bending deformation of the rail will be higher when the distance increases. The stresses in the
rail on the continuous support (CDM_Classic and CDM_Comfort) are much lower than in the rail on
discrete fixations.
The stresses are also lower in the rail foot than in the rail web. When it is a discrete fixations track (ballast,
direct_soft, direct_hard), the maximum stress in the web is located in the section halfway between the
fixations and the maximum stress in the rail foot is located on the edge of the fixation where the shear
stress is the highest. When it is a continuously supported rail (CDM_Classic, CDM_Confort), the
maximum stress in the rail web and the maximum stress in the rail foot are both located in the section
where the load is applied.
The stresses in the rail vary in function of the type of rail. The rails can be classified in function of their
maximum stress (from highest to lowest):
Ri59N -> NP4am -> 35G -> Ri53N -> EB50T for the tramway;
UIC50 -> UIC54 for the light rail.
TIP5-CT-2006-031312 Page 154 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
5.2.1.5. Conclusion
The sensitivity analysis carried and presented in this report leads to several interesting conclusions:
conclusions on the stresses in the rail and conclusions on the vibrations levels under the track.
The wheel/rail contact forces calculated in Vi-Rail for different combination of vehicle, rail type, support,
type and velocity were applied to track Finite Element track models in Nastran. The resulting stresses in
the rail are discussed in this report.
From the results, it can be seen that, the more the railpad is stiff, the highest the stress. The stress is much
lower when the rail is continuously supported. The stress also increases with the distance between the
rail fixation. The stresses in the rail web in curve are much higher than on straight track. This cannot be
said for the stresses in the rail foot. In curve, the stresses increases with the vehicle speed while it is not
the case in straight line.
The displacements of the rail and of the base under the railpads are computed in Vi-Rail. The rail
irregularities included in the track model (measured rail irregularities and step function rail irregularity)
causes these displacements to vary around an average value (the static deformation): vibrations occur.
These displacement in function of time are transformed into displacement in function of the frequency
with a Fast Fourier Transform algorithm. The derivation of these displacement spectra lead to vibration
velocity spectra. Several conclusions can be drawn by looking at these spectra.
The vibration levels increase with the speed. For measured rail irregularities, this increase is mainly
visible at lower and higher frequencies. For a step function irregularity, the increase is of a factor 20 *
log10(V2/V1) in all frequency bands.
The vibrations under the ballasted track are lower than under the CDM-Classic, CDM-Comfort and
Direct-Soft where they are similar. The Direct-Hard system gives the highest vibration levels.
An interesting behaviour has been discovered in the spectra: the lateral movement of the bogies in the
curves also create peaks at low frequencies in the vibration spectra. These peaks appear above a certain
vehicle velocity and increase in amplitude and frequency as velocity increases.
Finally, this study allows many conclusions to be drawn from its results. These conclusions have been
discussed here. All these conclusions help establish functional requirements for urban railway tracks.
TIP5-CT-2006-031312 Page 155 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
5.2.2 Sensitivity analysis on metro track parameters by means of numerical simulation of
train-track dynamic interaction
5.2.2.1. Overview
In order to investigate functional requirements for metro track systems, PoliMi has provided numerical
tools and simulations for performing parametric studies on track parameters. An extensive sensitivity
analysis has been performed, by means of numerical simulation of train-track dynamic interaction.
As agreed among SP5 partners, the Metro Madrid case study was chosen for PoliMis numerical
simulations. Since the focus is the metro track, vehicle characteristics were fixed, and variable track
parameters were considered.
The final objective of the sensitivity analysis was to evaluate the effects on track stability and vibration
mitigation of the following parameters:
different track design options (embedded rail, direct fixation, floating slab track) and rail support
stiffness values;
support stiffness variations corresponding to new/degraded track conditions;
axle load variations (tare and full load);
track irregularity corresponding to different levels of degraded track geometry;
rail profile and gauge.
5.2.2.2. Numerical simulation of train-track interaction
The numerical simulation of the dynamic interaction between train and track requires combined
modelling of the two subsystems involved. To this end, a numerical procedure has been developed at
Politecnico di Milano - Department of Mechanical Engineering, which is based on a finite element
schematisation of the track/structure and on a multi-body schematisation of the train. The two
subsystems are described through two separate sets of differential equations which are simultaneously
integrated in the time domain. The co-simulation procedure is based on the coupling of the train and the
track dynamics, as the result of the contact forces exchanged at wheel-rail interface: as it is shown in
Figure 2.2.1, these contact forces are a function of both the vehicle and the infrastructure state variables.
Both vertical and lateral dynamics of the entire system are taken into account.
Figure 5.2.29
TIP5-CT-2006-031312 Page 156 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
The numerical model for the simulation of train-structure interaction: vehicle and structure as separate
but interacting subsystems.
The following equations hold for the structure:
| | | | | | ( ) ( )
1 1
, , , , , , , ,
sv ss s s s s s v v vn vn s s s s s
M x C x K x F x x x x x x t F x x t + + = +
where [Ms], [Cs] and [Ks] are the mass, damping and stiffness matrices of the structure,
sv
F is the column
matrix of the generalised wheel-rail contact forces (which are a function of both the train and the
structure state coordinates),
ss
F represents the generalised terms due to the non linear internal forces
due to the particular formulation of some track components.
In the train multi-body model, each railcar or locomotive is composed of rigid bodies, connected by
means of linear/nonlinear elastic and damping elements that reproduce the primary and secondary
suspensions. The vehicle equations of motion are written with respect to an auxiliary moving frame of
reference, travelling along the ideal track centreline and following the carbody centre of gravity. The
vehicle relative motion with respect to this moving reference is considered.
For the generic i-th vehicle in the train assembly (n being the total number of cars), the following
equations hold:
| | | | | | ( ) ( )
1 1
, , , , , , , ,
vi n i vi vi vi s s v v vn vn vi vi vi vi vi
M x C x K x F x x x x x x t F x x t + + = +


where [Mvi], [Cvi], [Kvi] are the mass, damping and stiffness matrices of the i-th vehicle,
vi
F is the column
matrix containing the generalised wheel-rail contact forces and
n i
F

accounts for nonlinear suspension
components of the rail vehicle.
The dependence on time t of the generalised contact forces is due to track and wheel irregularity which,
for a given train speed, are assigned functions of time.
A key element in the simulation code is the non-linear wheel-rail contact model, which allows to
accurately reproduce the normal and tangential contact forces exchanged at wheel rail interface.
The normal forces are evaluated through a multi-Hertzian model, which allows either to take into
account the presence of multiple contact points, e.g. when contacts occur both on the tread and on the
flange, or to approximate complex non-hertzian contact patches by one or more elliptic patches.
The lateral and longitudinal contact forces are then determined through the heuristic ShenHedrick
Elkins contact model, starting from the longitudinal and lateral creepages on each contact area, which are
derived from the wheelset kinematics.
TIP5-CT-2006-031312 Page 157 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
(a) VEHICLE MODEL
As agreed among SP5 partners, the Metro Madrid case study was chosen for PoliMis numerical
simulations. In order to investigate the functional requirements for metro track infrastructure, vehicle
characteristics were fixed, and a sensitivity analysis on track parameters was performed.
Figure 5.2.30 shows the vehicle considered in the simulations. The main vehicle characteristics are
reported in table 5.2.11.
Figure 5.2.30 The Metro Madrid train
Figure 5.2.31 The Metro Madrid train
maximum axle load 154 kN
tare axle load 96 kN
wheelbase 2.2 m
pivot pitch 11.1 m
Table 5.2.11 Main characteristics of the Metro Madrid vehicle
The multibody model of one Metro Madrid coach was implemented into PoliMis simulator (Fig. 5.2.32):
carbody and bogie frames are assumed to be rigid;
wheelsets are modelled as flexible bodies, according to modal superposition approach;
elastic and damping connection elements reproduce the primary and secondary suspensions.
TIP5-CT-2006-031312 Page 158 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
Figure 5.2.32 Vehicle multi-body model
(b) TRACK MODEL
For the purpose of this research, a finite element model of the embedded-rail track system was developed
(figure 5.2.33) and implemented into PoliMis train-track dynamic simulation code.
A linear visco-elastic bed reacting in vertical, lateral and torsional direction reproduces the rubber jacket
enclosing the rail. Its visco-elastic characteristics were identified during the lab tests performed by PoliMi
in SP1, on small-scale REMS samples.
Both the rails and the tunnel are modelled as equivalent Euler-Bernoulli beams: while the bending
stiffness and the mass per unit length of the rail can be easily determined form the rail producer
catalogue, the properties of the equivalent bea m representing the tunnel and those of the visco-elastic
bed representing the ground were identified from impulsive tests carried out by PoliMi in Milano
underground some years ago. The visco-elastic bed representing the ground was introduced to allow
investigating the transmissibility (i.e. the capability of filtering out vibrations induced by train passage) of
the considered REMS tracks with respect to more traditional track systems.
Figure 5.2.33 Cross-section of the implemented finite element model of the embedded-rail track system
TIP5-CT-2006-031312 Page 159 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
As anticipated before, the visco-elastic characteristics of the elastic bed enclosing the rail were identified
during the lab tests performed by PoliMi in SP1, on small-scale REMS samples.
The test bench (figure 5.2.34) consists of a restraining frame, which carries the hydraulic actuator used to
apply the test loads to the samples. All the tests were performed in force control.
The force applied on top of the rail is measured by a load cell, while the motion of the rail with respect to
the concrete block (both being assumed to be rigid bodies) is measured by means of laser transducers.
Static and dynamic characterization tests were performed on all the three REMS samples (COMPACT,
CLASSIC and COMFORT), to assess their stiffness and damping characteristics as a function of
frequency, load amplitude and preload. The samples were tested by applying both a purely vertical force
and a 26deg inclined load (see figure 5.2.35).
Table 5.2.12 summarizes the measured stiffness parameters, for the three REMS samples. These data were
used in the train-track interaction simulations described in the next paragraphs.
Figure 5.2.34 Test bench for REMS samples characterization tests
TIP5-CT-2006-031312 Page 160 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
Figure 5.2.35 Test configurations: vertical and inclined load
vertical stiffness lateral stiffness
static dynamic static dynamic
(0 deg) (0.9Q0.1Q@10 Hz) (26 deg) (0.9Q0.1Q@10 Hz)
compact 78 1175 28 55
classic 44 105 22 50
comfort 14 50 20 45
Table 5.2.12 REMS stiffness parameters measured during PoliMis lab tests
TIP5-CT-2006-031312 Page 161 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
5.2.2.3. Simulations in tangent track
Tangent track numerical simulations of train-track dynamic interaction were first carried out to assess the
performance of metro track systems in terms of vibration mitigation.
A sensitivity analysis on the principal parameters involved was performed, considering:
different track design options (embedded rail, direct fixation, floating slab track) and rail support
stiffness values;
support stiffness variations corresponding to new/degraded track conditions;
axle load variations (tare and full load);
track irregularity corresponding to two different levels of degraded track geometry.
The sensitivity analysis concentrated on the embedded rail track system and the starting point were the
support stiffness values of the three classes of REMS track (table 5.2.2). The performance of the embedded
rail system was compared to that of direct fixation and floating slab solutions.
Both wheel and rail irregularity input data were considered in all the simulations. While measured
irregularity patterns were considered for the wheel, track irregularity input was generated according to
the PSD functions defined in the ORE B176 Standard.
(a) SIMULATIONS PERFORMED
A complete list of all the performed tangent track simulations is reported in table 5.2.13.
tare/full load track irregularity degraded #
ER comfort both 2 Y 8
ER classic both 2 Y 8
ER compact both 2 Y 8
DF both 2 N 4
FS both 2 N 4
total number of simulations 32
Table 5.2.13 List of the simulations for tangent track running (V = 110 km/h)
As an example, figure 2.3.1 shows the time histories of the calculated vertical rail and tunnel accelerations
during tangent track running at 110 km/h for the three REMS stiffness classes. These simulations refer to
nominal (new) track parameters and to vehicle in full-load condition.
TIP5-CT-2006-031312 Page 162 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
-80
-60
-40
-20
0
20
40
60
80
s
m
/
s
2
COMFORT - FULL LOAD
rail
tunnel
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
-80
-60
-40
-20
0
20
40
60
80
s
m
/
s
2
CLASSIC - FULL LOAD
rail
tunnel
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
-80
-60
-40
-20
0
20
40
60
80
s
m
/
s
2
COMPACT - FULL LOAD
rail
tunnel
Figure 5.2.36 Numerical simulations for the estimation of vibration mitigation performance (REMS). Metro
Madrid train running in tangent track at 110 km/h (full load). Simulations output: rail and tunnel
acceleration time histories.
It can be clearly seen that, when increasing the track stiffness, the maximum rail vibration reduces, while
higher vibrations are transmitted to the tunnel. A significant change in terms of duration can also be
observed.
TIP5-CT-2006-031312 Page 163 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
(b) TRACK DYNAMIC PERFORMANCE ESTIMATION: VIBRATION MITIGATION
The time-domain numerical simulations of train-track interaction in tangent track (see table 5.2.3) were
post-processed, so as to obtain summarizing information concerning the track dynamic performance in
the different conditions. As an example, figures 5.2.37 and 5.2.38 show the rail and tunnel one-third
octave acceleration spectra (reference acceleration 1.0e-6 m/s
2
) and the corresponding vibration
attenuation, for the three stiffness classes of the embedded rail track and the direct fixation track. In all
the simulations, the Metro Madrid train is running in tangent track at 110 km/h, in full load condition.
As expected, among the three stiffness classes of the embedded rail track, the best performance in terms
of vibration mitigation is that of the COMFORT solution, which proves to be better than the direct
fixation track, at any frequency above 40 Hz.
10
1
10
2
10
3
60
70
80
90
100
110
120
130
140
Hz
d
B
RAIL ACCELERATION - FULL LOAD
ER comfort
ER classic
ER compact
DF
10
1
10
2
10
3
60
70
80
90
100
110
120
130
140
Hz
d
B
TUNNEL ACCELERATION - FULL LOAD
ER comfort
ER classic
ER compact
DF
Figure 5.2.37 Metro Madrid train running in tangent track at 110 km/h (full load): rail and tunnel one-third
octave acceleration spectra (3 stiffness classes of embedded rail vs. direct fixation)
10
1
10
2
10
3
0
5
10
15
20
25
30
35
40
45
50
Hz
d
B
VIBRATION ATTENUATION - FULL LOAD
ER comfort
ER classic
ER compact
DF
Figure 5.2.38 Metro Madrid train running in tangent track at 110 km/h (full load): vibration attenuation from rail
to tunnel (3 stiffness classes of embedded rail vs. direct fixation)
TIP5-CT-2006-031312 Page 164 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
(c) EFFECT OF TRACK DEGRADATION
Figures 2.3.4, 2.3.5 and 2.3.6 show how the performance of the three considered classes of the embedded-
rail track varies, as a result of track degradation. The effect of track degradation due to repeated load
cycles was considered by increasing the stiffness of each one of the three tracks by 20%.
Figures 2.3.4, 2.3.5 and 2.3.6 show the rail and tunnel one-third octave acceleration spectra (1.0e-6m/s
2
reference acceleration), and the corresponding vibration attenuation, in case of new and degraded track,
for the 3 stiffness classes of the embedded rail track. In all the simulations, the Metro Madrid train is
running in tangent track at 110 km/h, in full load condition. For all the three stiffness classes, no
significant variation in vibration mitigation performance can be observed.
10
1
10
2
10
3
60
70
80
90
100
110
120
130
140
Hz
d
B
rail acceleration
tunnel acceleration
tunnel acceleration (degr.track)
rail acceleration (degr.track)
10
1
10
2
10
3
0
5
10
15
20
25
30
35
40
45
50
Hz
d
B
vibration attenuation
vibration attenuation (degr.track)
Figure 5.2.39 Metro Madrid train running in tangent track at 110 km/h (full load): rail and tunnel one-third
octave acceleration spectra and corresponding vibration attenuation, in case of new (nominal
stiffness) and degraded (+20% stiffness) REMS COMPACT track
TIP5-CT-2006-031312 Page 165 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
10
1
10
2
10
3
60
70
80
90
100
110
120
130
140
Hz
d
B
rail acceleration
tunnel acceleration
tunnel acceleration (degr. track)
rail acceleration (degr. track)
10
1
10
2
10
3
0
5
10
15
20
25
30
35
40
45
50
Hz
d
B
vibration attenuation
vibration attenuation (degr. track)
Figure 5.2.40 Metro Madrid train running in tangent track at 110 km/h (full load): rail and tunnel one-third
octave acceleration spectra and corresponding vibration attenuation, in case of new (nominal
stiffness) and degraded (+20% stiffness) REMS CLASSIC track
TIP5-CT-2006-031312 Page 166 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
10
1
10
2
10
3
60
70
80
90
100
110
120
130
140
Hz
d
B
rail acceleration
tunnel acceleration
tunnel acceleration (degr. track)
rail acceleration (degr. track)
10
1
10
2
10
3
0
5
10
15
20
25
30
35
40
45
50
Hz
d
B
vibration attenuation
vibration attenuation (degr. track)
Figure 5.2.41 Metro Madrid train running in tangent track at 110 km/h (full load): rail and tunnel one-third
octave acceleration spectra and corresponding vibration attenuation, in case of new (nominal
stiffness) and degraded (+20% stiffness) REMS COMFORT track
TIP5-CT-2006-031312 Page 167 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
5.2.2.4. Simulations in curve
As for tangent track running, curved track numerical simulations were carried out to assess the
performances of metro track systems in terms of track stability. The following parameters were taken into
account for the sensitivity analysis:
different track design options (embedded rail, direct fixation, floating slab track) and rail support
stiffness values;
support stiffness variations corresponding to new/degraded track conditions;
axle load variations (tare and full load);
The sensitivity analysis concentrated on the embedded rail track system and the starting point were the
support stiffness values of the three classes of REMS track (table 5.2.2). The performance of the embedded
rail system was compared to that of direct fixation and floating slab solutions. Both wheel and rail
irregularity input data were considered in all the simulations.
(a) SIMULATIONS PERFORMED
Track stability was analysed in terms of rail vertical deflection and gauge widening. A reference curve of
300 m radius and 150 mm of cant was taken into account and the vehicle speed was imposed to be equal
to 80 km/h. Simulations were performed for both tare and fully loaded vehicle and with new and worn
wheel/rail profiles. For the comfort embedded track, the influence of degradation of track characteristics
was also investigated. Degraded characteristics are taken from experimental tests carried out at
Politecnico di Milano. Table 5.2.14 sums up the simulations carried out.
tare/full load W/R profiles degraded #
ER comfort both 2 Y 8
ER classic both 2 Y 4
ER compact both 2 Y 4
DF both 2 N 4
FS both 2 N 4
total number of simulations 24
Table 5.2.14 List of the simulations for curved track running (R = 300 m, h = 150 mm, V = 80 km/h)
(b) TRACK DYNAMIC PERFORMANCE ESTIMATION: TRACK STABILITY
As an example, figure 2.4.1 shows the achieved results in terms of time histories of vertical rail deflection
(of both inner and outer rail) and gauge widening at a measuring point 8 m away from the end of the
transition curve of the considered reference curve for the various track types considered. Both tare and
fully loaded vehicle conditions are shown.
TIP5-CT-2006-031312 Page 168 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
Vertical rail deflection - R= 300 m, Cant = 150 mm, V = 80 km/h, Tare load (axle load = 96 kN)
[s]
[m
m
]
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
0
0.2
0.4
0.6
0.8
1
Gauge widening - R = 300 m, Cant = 150 mm, V = 80 km/h, Tare load (axle load = 96 kN)
[s]
[m
m
]
inner rail - 20m out of transition
outer rail - 20mout of transition
20mout of transition
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
Vertical rail deflection - R= 300 m, Cant = 150 mm, V = 80 km/h, Full load (axle load =154 kN)
[s]
[m
m
]
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
0
0.2
0.4
0.6
0.8
1
Gauge widening - R= 300 m, Cant = 150 mm, V = 80 km/h, Full load (axle load =154 kN)
[s]
[m
m
]
inner rail - 20m out of transition
outer rail - 20mout of transition
20mout of transition
COMPACT tare load COMPACT full load
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
Vertical rail deflection - R= 300 m, Cant = 150 mm, V = 80 km/h, Tare load (axle load = 96 kN)
[s]
[m
m
]
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
0
0.2
0.4
0.6
0.8
1
Gauge widening - R = 300 m, Cant = 150 mm, V = 80 km/h, Tare load (axle load = 96 kN)
[s]
[m
m
]
inner rail - 20m out of transition
outer rail - 20mout of transition
20mout of transition
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
Vertical rail deflection - R= 300 m, Cant = 150 mm, V = 80 km/h, Full load (axle load =154 kN)
[s]
[m
m
]
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
0
0.2
0.4
0.6
0.8
1
Gauge widening - R= 300 m, Cant = 150 mm, V = 80 km/h, Full load (axle load =154 kN)
[s]
[m
m
]
inner rail - 20m out of transition
outer rail - 20mout of transition
20mout of transition
CLASSIC tare load CLASSIC full load
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
Vertical rail deflection - R= 300 m, Cant = 150 mm, V = 80 km/h, Tare load (axle load = 96 kN)
[s]
[m
m
]
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
0
0.2
0.4
0.6
0.8
1
Gauge widening - R = 300 m, Cant = 150 mm, V = 80 km/h, Tare load (axle load = 96 kN)
[s]
[m
m
]
inner rail - 20m out of transition
outer rail - 20mout of transition
20mout of transition
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
Vertical rail deflection - R= 300 m, Cant = 150 mm, V = 80 km/h, Full load (axle load =154 kN)
[s]
[m
m
]
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
0
0.2
0.4
0.6
0.8
1
Gauge widening - R= 300 m, Cant = 150 mm, V = 80 km/h, Full load (axle load =154 kN)
[s]
[m
m
]
20mout of transition
inner rail - 20m out of transition
outer rail - 20mout of transition
COMFORT tare load COMFORT full load
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
Vertical rail deflection - R = 300 m, Cant = 150 mm, V = 80 km/h, Tare load (axle load = 96 kN)
[s]
[m
m
]
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
0
0.2
0.4
0.6
0.8
1
Gauge widening - R = 300 m, Cant = 150 mm, V = 80 km/h, Tare load (axle load = 96 kN)
[s]
[m
m
]
20mout of transition
inner rail - 20mout of transition
outer rail - 20mout of transition
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
Vertical rail deflection - R= 300 m, Cant = 150 mm, V = 80 km/h, Full load (axle load =154 kN)
[s]
[m
m
]
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4
0
0.2
0.4
0.6
0.8
1
Gauge widening - R= 300 m, Cant = 150 mm, V = 80 km/h, Full load (axle load =154 kN)
[s]
[m
m
]
20mout of transition
inner rail - 20m out of transition
outer rail - 20mout of transition
Direct Fixation tare load Direct Fixation full load
Figure 5.2.42 Metro Madrid train running on a curved track (300 m radius, 150 mm cant, 80 km/h speed):
Vertical rail deflection (upper graphs) and gauge widening (lower graph) of the various tracks
considered for both tare and full load
TIP5-CT-2006-031312 Page 169 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
The passages of the four wheelsets (two bogies) are clearly visible in the vertical rail deflection graph. The
difference between the passage of the first bogie and the second one is due to the roll motion of the
carbody that is still present at that position. It can be clearly seen that, increasing the vehicle load (form
an axle load of 96kN in tare conditions to and axle load of 154kN in fully loaded conditions), the vertical
rail deflection increases. Moreover, for what concerns embedded tracks, vertical rail deflection increases
from the stiffest one (compact track) to the most compliant one (comfort track) although no big
differences between compact and classic track can be observed. Direct fixation track, instead has a higher
vertical stiffness than embedded track thus leading to a vertical rail deflection that is almost half of that of
compact track.
The time history of gauge widening shows the passage of the four wheelsets (more evident for the fully
loaded vehicle) with superimposed oscillations, especially for the direct fixation track that has discrete
supports and lower damping, generated by the train passage that propagate along the rails. The highest
peak is reached for the first wheelset that has the highest angle of attack and therefore generates the
biggest lateral force. The gauge widening dies out less rapidly than the vertical rail deflection thus
leading to a residual gauge widening between the passage of the two bogies. For what concerns
embedded tracks, the increased deformability form compact to comfort track determines and increased
gauge widening (from 0.45 mm for the compact track in tare load conditions to 0.83 mm for the comfort
track in tare load conditions). Moreover, it should be observed that the direct fixation track, due to its
smaller damping characteristics, "feels" the train passage earlier than embedded tracks and that it
experiences a higher peak gauge widening due to the fact that discrete supports are less capable to re-
distribute lateral forces
The influence of track degradation and wear of wheel/rail profiles is, as expected negligible on vertical
rail deflection and gauge widening. Table 5.2.15 sums up the results. For what concerns the vertical rail
displacement, reference to the outer rail has been made being this rail the most stressed one due to the
load transfer. For what concerns the gauge widening, instead, reference to the passage of the first
wheelset has been made being this the peak value as explained above.
Tare load Full load
Vertical rail
deflection
Gauge widening Vertical rail
deflection
Gauge widening
[mm] [mm] [mm] [mm]
COMPACT 0.577 0.448 0.925 0.670
CLASSIC 0.701 0.603 1.114 0.761
COMFORT 1.773 0.625 2.904 1.131
Direct Fixation 0.324 0.832 0.413 1.181
Table 5.2.15 Maximum values for the different track considered at different loads
(R = 300 m, h = 150 mm, V = 80 km/h)
It can be clearly seen that the direct fixation track behaves better than embedded rail systems for what
concerns the vertical rail deflection while it behaves worse for what concerns the gauge widening. This is
due to the fact that, for embedded rail systems, the rail is laterally supported by continuous rubber
elements. Moreover, embedded track suffer less than direct fixation track from corrugation phenomena
due to the presence of a continuous support.
TIP5-CT-2006-031312 Page 170 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
5.2.2.5. Numerical-experimental comparison
A joint test campaign on both direct fixation and comfort embedded track was performed on Metro
Madrid line. Results were used to validate the numerical simulations with particular reference to vertical
rail displacements and gauge widening. The experimental set-up was composed by a series of
displacement transducers to detect vertical and lateral movements of particular rail sections: for the direct
fixation track, a straight track was instrumented while for the embedded track system a curve having
300 m radius was instrumented. As an example figure 5.2.43 shows a picture of the instrumented direct
fixation track section while figure 5.2.44 shows a picture of the instrumented embedded track section.
Figure 5.2.43 Instrumented test section of the direct fixation track
Figure 5.2.44 Instrumented test section of the embedded track
Tests were performed using a standard Metro Madrid coach passing on the instrumented sections at
different speeds in both directions. As an example, figures 5.2.45 5.2.46 show acquired channels for a
vehicle passage at about 80 km/h: figures 5.2.45 and 5.2.47 show the time-histories of the vertical rail
TIP5-CT-2006-031312 Page 171 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
deflection (the blue line is referred to the outer rail and the magenta line to the inner rail) while figures
5.2.46 and 5.2.48 show the time histories of the rail lateral deflections (the blue line is referred to the outer
rail and the red line to the inner rail) and gauge widening (green line) for two different passages (to and
forth). It can be observed that the maximum measured vertical rail deflection is about 2 mm for both
passages while the maximum measured gauge widening is about 0.6 mm when running in one direction
and 1.3 mm when running the opposite direction. This difference is due to the fact that test section was
close to the transition curve. Thus, in one direction the value is associated to full curve running while in
the other direction it is influenced by transient loads due to the transition from curve entrance to full
curve.
Both the measured maximum vertical rail deflection and the gauge widening values for full curve
running are very close to the corresponding values simulated through the numerical train-track
interaction model (paragraph 4) that correctly accounts for the deformability of the rubber elements and
precisely reproduces wheel rail contact conditions.
-3,0
-2,5
-2,0
-1,5
-1,0
-0,5
0,0
0,5
1,0
3 4 5 6 7 8
D
e
s
p
l
a
z
a
m
i
e
n
t
o
(
m
m
)
Tiempo (seg.)
DEFLEXIN HILO EXTERIOR DEFLEXIN HILO INTERIOR
Figure 5.2.45 Metro Madrid train running on a curved track (300 m radius, 150 mm cant, 80 km/h speed, full
curve): time-history of the measured vertical deflections of both inner (magenta line) and outer (blue
line) rails
TIP5-CT-2006-031312 Page 172 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
-1,3
-0,6
0,0
0,6
1,3
3 4 5 6 7 8
D
e
s
p
l
a
z
a
m
i
e
n
t
o
(
m
m
)
Tiempo (seg.)
P41 CABEZA HILO EXTERIOR P31CABEZA HILO INTERIOR ANCHOVA
Figure 5.2.46 Metro Madrid train running on a curved track (300 m radius, 150 mm cant, 80 km/h speed, full
curve): time-history of the lateral deflections of both inner (blue line) and outer (red line) rails and
gauge widening (green line)
-3,0
-2,5
-2,0
-1,5
-1,0
-0,5
0,0
0,5
1,0
6 7 8 9 10 11 12
D
e
s
p
l
a
z
a
m
i
e
n
t
o
(
m
m
)
Tiempo (seg.)
DEFLEXIN HILO EXTERIOR DEFLEXIN HILO INTERIOR
Figure 5.2.47 Metro Madrid train running on a curved track (300 m radius, 150 mm cant, 80 km/h speed, end of
transition curve): time-history of the measured vertical deflections of both inner (magenta line) and
outer (blue line) rails
TIP5-CT-2006-031312 Page 173 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
-1,3
-0,6
0,0
0,6
1,3
1,9
7 8 9 10 11 12 D
e
s
p
l
a
z
a
m
i
e
n
t
o
(
m
m
)
Tiempo (seg.)
P41CABEZA HILO EXTERIOR P31 CABEZA HILO INTERIOR ANCHO VA
Figure 5.2.48 Metro Madrid train running on a curved track (300 m radius, 150 mm cant, 80 km/h speed, end of
transition curve): time-history of the lateral deflections of both inner (blue line) and outer (red line)
rails and gauge widening (green line)
TIP5-CT-2006-031312 Page 174 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
5.2.2.6. Concluding remarks
A numerical train track interaction model was set-up to carry out a sensitivity analysis on metro track
design parameters to evaluate their effects on track stability and vibration transmission. The following
parameters were considered:
different track design options (embedded rail, direct fixation, floating slab track) and rail support
stiffness values;
support stiffness variations corresponding to new/degraded track conditions;
axle load variations (tare and full load);
track irregularity corresponding to two different levels of degraded track geometry.
For straight track running, it is shown that increasing the track stiffness the maximum rail vibration
reduces, while higher vibrations are transmitted to the tunnel. A significant change in terms of duration
of the vibration can also be observed. Among the three stiffness classes of the embedded rail track
system, the best performance in terms of vibration mitigation is that of the comfort solution that proves to
be better than the direct fixation track, at any frequency above 40 Hz. For what concerns track
degradation, no significant variation in vibration mitigation performance can be observed for all the three
stiffness classes considered.
For curved track running, it is shown that the direct fixation track behaves better than embedded rail
systems for what concerns the vertical rail deflection while it behaves worse for what concerns the gauge
widening.
The numerical experimental comparison shows that both the measured maximum vertical rail
deflection and the gauge widening values for full curve running are reproduced by the numerical train-
track interaction model that correctly accounts for the deformability of the rubber elements and precisely
reproduces wheel rail contact conditions.
TIP5-CT-2006-031312 Page 175 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
5.2.3 Light rail parameters based on national guidelines and recommendations
5.2.3.1. Introduction:
The BOStrab (=German federal Regulations on the Construction and Operation of Light rail Transit
Systems) is a national guideline for light rail. It is divided in the following eight parts:
1 General
2 Operating Panel
3 Operating Staff
4: Operational Installations
5: Vehicles
6 Operations
7 Procedural Formalities
8 Non-compliances, conclusion and transitional arrangements
Because of the long experience (the first issue was published 1938) in light rail behind the BOStrab
especially the fourth part (operational installations) and the BOStrab based recommendations often were
adapted as guidelines e.g. for Poland, Finland, Austria, Norway and partially Spain.
The Rules on the Alignment of Rail Systems for light rail contain essential parameters for Light Rail
Installations in particular. In consideration of these rules, maintenance costs could be reduced.
Below, an example for a possible guideline summarising the basic data of the recommendations.
5.2.3.2. Rules on the Alignment of Rail Systems
(a) GENERAL OBJECTIVES
The objective of the alignment is the lining-up and overlapping of the alignment elements in a way
favourable to the dynamics of vehicle movements as well as optimisation of the alignment parameters in
respect of
Safety,
Speed,
Dynamics of vehicle movements and ride comfort,
Economic efficiency of the structure and operation including the maintenance.
These objectives are reached by applying the standard values.
(b) STANDARD ANDLIMIT VALUES
The alignment parameters are to be determined on the basis of the project speed vp and the standard
values of these Rules. If deviations are made from the standard values for compelling reasons, the margin
between the standard values and the limit values of these Rules can be applied provided that the project
speed is observed. The limit values (as minimum or maximum values, respectively) of the alignment
parameters shall not be exceeded.
TIP5-CT-2006-031312 Page 176 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
(c) SPEEDS
A uniform project speed vp is to be determined for the alignment of a route network or for parts of the
route network. This speed is to be adjusted to the maximum speed of the present and future vehicles. The
project speed shall not be less than vp = 50 km/h for in-street and segregated track formations and not
less than vp = 70 km/h for independent track formations. Lower speeds can be assumed for those parts of
the network on which there is no regular passenger transport. If it is necessary to deviate from the
standard values of the alignment at constraints, the permissible speed per v determined by way of the
limit values of the alignment parameters shall correspond at least to the project speed vp. If the limit
values of the alignment parameters cannot be observed with the project speed vp in special
circumstances, a lower permissible speed per v is to be determined in deviation to the project speed.
(d) STRAIGHT
The track shall be as straight as at all possible. A straight between two curves shall have the length
| | m
v
l
p
g
10
> (eq 1)
It shall not be shorter than the distance between the centres of the running gears of the vehicles mainly
operated and not smaller than 6 m. If this length cannot be observed, the intermediate straight shall not
be considered by the calculation of the permissible speed per v.
(e) CURVE AND SUPERELEVATION
(i) Transverse Acceleration
For the transverse acceleration in the curve
(

=
153 * 6 . 3
2
s
m u
r
v
a
q
(eq 2)
applies.
As a rule, the alignment is based on a slight, positive transverse acceleration (aq ~ 0.2 m/s)).
The transverse acceleration shall not be higher than aq= 0.65 m/s. In special circumstances he limit value
aq = 0.98 m/s is permitted; this value is also decisive for the proof of the permissible speed
(

+
=
h
km u r
perv
8 . 11
) 150 ( *
(eq 3)
(ii) Curve Radius
For the curve radius
| | m
a u
v
r
q
* 153
* 8 . 11
+
= (eq 4)
TIP5-CT-2006-031312 Page 177 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
follows from (eq. 2).
Curve radii are to be as big as possible as a function of the project speed vp.
On independent track formations curve radii of route tracks are to be so big that the maximum line speed
max v is not restricted in case of a transverse acceleration of aq = 0.65 m/s. However, they shall not be
smaller than r = 240 m. On segregated and in-street track formations, the relevant local conditions are to
be considered in view of these Rules. The minimum radius min r amounts to 25 m. If track sections are
connected via an intermediate curve, this curve shall have the length of an intermediate straight in
accordance with Section 5. If possible, the platforms shall be located at a straight track. If platforms have
to be located in a curve, it is usually not possible to observe the requirements for handicapped-friendly
boarding and alighting.
(iii) Superelevation
For the partial compensation of the transverse acceleration the outside running rail is laid higher in the
curve than the inside running rail by the dimension of the superelevation u.
| | m a
r
v
u
q
* 153
* 8 . 11
2
= (eq 5)
The speed decisive for the superelevation to be realised is to be determined on the basis of the
speed/distance chart in accordance with Section 3.5.
The compensating superelevation uo amounts to
] [
* 8 . 11
0
mm
r
v
u = (eq 6)
However, as a rule the superelevation is to be so selected that there is a slight, positive transverse
acceleration aq ~ 0.2 m/s (normal super-elevation nor u).
| | mm
r
v
noru 30
* 8 . 11
= (eq 7)
During a transverse acceleration of aq = 0.65 m/s the superelevation shall not be smaller than
] [ 100
* 8 . 11
mm
r
v
u = (eq 8)
In special circumstances the minimum superelevation during a transverse acceleration of aq = 0.98 m/s
amounts to
] [ 150
* 8 . 11
min mm
r
v
u = (eq 9)
The superelevation shall not be bigger than u = 150 mm; in special circumstances a value up to the
maximum superelevation max u = 165 mm is permitted.
Platform tracks shall be designed without superelevation.
TIP5-CT-2006-031312 Page 178 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
(f) TRANSITIONCURVE AND SUPERELEVATIONRAMP
(i) Transition Curve
As a rule, there has to be a transition curve between a straight and a curve as well as between curves of
different curvatures to minimise the transverse jerk.
For the transverse jerk C in the transition curve
(

A
=
* 6 . 3
*
s
m
l
a v
C
u
q
(eq 10)
applies.
aq is the difference in transverse acceleration between the beginning and the end of the transition curve.
The transverse jerk shall not be bigger than max C = 0.67 m/s.
The minimum length of the transition curve amounts to
] [
4 . 2
*
min m
a v
l
q e
u
A
= (eq 11)
due to the transverse jerk. The transition curve and the superelevation ramp shall have the same length as
well as the same beginning and the same end (see Section 7.2).
As the shape of the transition curve the Clothoid
] [ * m l r A
U
= is to be
preferred.
If it is not possible to realise a transition curve, the permissible speed amounts to
(

A
=
h
km
k
l C perv
v
3
1000
* * max * 6 . 3 (eq 13)
In this case the transverse jerk has an effect on a virtual length lv. This length depends on the vehicle and
is influenced by e.g. the distance between the centres of the running gears, the distance between the axles
in the running gear and the rigidity of the transverse springs.
For I
v
the distance between the centres of the running gears of the vehicle type mainly operated is to be
selected. Simplified, lv = 6 m can be applied.
The difference in curvature amounts to
.
1000 1000
2 1
1 2
r r with
r r
k > = A
The curvatures for opposite curves are to be added and the curvatures for curves of the same kind to be
subtracted.
TIP5-CT-2006-031312 Page 179 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
For the transition without a transition curve between two curves of different curvatures the permissible
speed per v results to
(

A
=
h
km
k
l
perv
v
3
* 5 . 31 (eq 13a)
with C = 0.67 m/s
3
according to eq. 13.
With l
v
= 6 m it follows that
(

A
=
h
km
k
perv
3
1
* 7 . 5 (eq 13b)
For the transition without a transition curve between a straight and a curve the permissible speed per v
results to
(

=
h
km
r l perv
v
3
* * 15 . 3 (eq 13c)
with C = 0.67 m/s according to eq. 13.
With l
v
= 6 m it follows that
(

=
h
km
r perv
3
* 7 . 5 (eq 13d)
The permissible speed per v resulting from eq. 13 can be rounded up to full 5 km/h, except if the radii are
smaller than 140 m.
(ii) Superelevation Ramp
The transition between a track section not superelevated and a superelevated section or between two
track sections with different superelevation is realised by way of a superelevation ramp.
Usually, such a superelevation ramp is straight.
The standard gradient of the superelevation ramp amounts to
p
v m * 10
1 1
= (eq 14)
In special circumstances the biggest gradient of the superelevation ramp amounts to
perv m * 6
1 1
= , however, maximum
300
1
(eq15)
On the basis of the gradient of the superelevation ramp the length of the ramp amounts to
] [
1000
*
m
u m
l
R
= (eq 16)
TIP5-CT-2006-031312 Page 180 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
If the superelevation ramp and the transition ramp have different lengths, the longer length of the two
lengths lR and lU shall be used as the joint length with the same beginning and the same end.
If the superelevation ramp is located partially in a curve, there has to be superelevation according to eq. 9
in each point of the transition curve.
Between two superelevation ramps with different gradients there has to be a track section without
superelevation or with constant superelevation. This section has to have the minimum length of an
intermediate straight in accordance with Section 5. It is allowed to do without this intermediate length if
the gradient of the two superelevation ramps against one another does not exceed the gradient according
to eq. 15.
If opposite curves follow one after the other, the superelevation of the first curve shall continuously be
transferred to the superelevation of the second curve (track scissors).
(g) GRADIENT AND CHANGE INGRADIENT
(i) Gradient
The gradient of the route tracks is to be in agreement with the starting and braking power of the vehicles.
As a rule, the gradient shall not exceed the value I = 40 .
In case of difficult topographical conditions, higher gradients can be realised if the vehicles are designed
accordingly.
Tracks at stops/stations, terminal loops and depots shall not have a gradient. The gradient at stops of in-
street rail systems shall only exceed 40 in special circumstances.
In special circumstances, there can be gradients in terminal loops and depots if it is prevented that the
trains can roll away.
(ii) Transition from One Gradient to Another
(1) As a rule, changes in the gradient shall be rounded with a radius of
2
* 4 . 0
p a
v r > (eq 17)
however at least ra = 1000 m. In special circumstances
2
* 25 . 0 perv r
a
> (eq 18)
however at least ra = 625 m, is permitted. Additional sighting distance checks are not affected.
There shall not be a change in the gradient in superelevation ramps. If this cannot be avoided, the
transition is to amount to
m
v
r
p
a
3
* 6
> (eq 19)
however at least ra = 2000 m.
TIP5-CT-2006-031312 Page 181 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
(3) In case of in-street track formations the transition from one gradient to another is to be adapted to the
local conditions; the minimum transitions that can be passed by the vehicles are to be observed.
(4) The standard transition according to eq. 17 also applies to switches in valleys, but only to switches on
hilltops if it can be ensured that the switch blade or the movable crossing lies on the slide chair within the
area of the point; otherwise the transition radius shall amount to at least ra = 5000 m or this switch area
shall be at least 5 m away from the vertical graduated transition curve.
(h) DEVIATIONS FOR NON-STANDARDGAUGE RAIL SYSTEMS
Besides the standard gauge (1435 mm) other gauges are included in the scope of these Rules. The gauges
deviating slightly from the standard gauge are considered to be standard gauges. The below table shows
the deviations from the formulas and numerical values mentioned in Sections 6 to 8 that apply to the
gauges s = 1100 mm and s = 1000 mm (metre gauge).
Table 5.2.16
TIP5-CT-2006-031312 Page 182 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
5.2.4 Applicability of rail standards to urban rail (tram & metro)
Appendix A5.2.3 (separate xls-file) contains an overview on the existing CEN and CENELEC standards
evaluated for their applicability.
5.2.5 Overview of a new embedded track standard under development in the US by
APTA/AREMA
5.2.5.1. Need for recommended practices
WP5.2 is tasked with the development of performance specifications for embedded track systems.
In early 2007, the APTA Track Noise & Vibrations subcommittee came to the conclusion that the Track
design handbook for light rail transit (TCRP Report 57) was no longer up to date and the AREMA
Manual of Railway Engineering was not up to date and not complete in the Chapters covering Rail
Transit issues.
5.2.5.2. Five priorities
The APTA and AREMA committees defined five priorities:
Non-ballasted track;
Embedded (paved) track;
Corrosion control;
Wheel/Rail interface issues;
Sharp curve design.
5.2.5.3. Work plan
(a) ISSUES TO BE ADDRESSED
The updated documents must incorporate the following issues:
Use of Tee rail vs. Grooved/Girder rail;
Slab design: reinforced or non-reinforced;
Optimum track modulus;
Method(s) of controlling stray currents;
Optimum track gauge to wheel gauge clearance;
Method(s) of fastening rails;
Curve guarding methods;
Controlling noise and vibrations.
(b) KNOWLEDGE INVENTORY
It was established that most items could be addressed with knowledge readily available within the
group. For others research statements were developed for submission to TRB (Transportation Research
Board) and TCRP (Transport Cooperative Research Program).
TIP5-CT-2006-031312 Page 183 of 289
URBAN TRACK Issued: November 11, 2010
D0602_M48_UITP_SP5.doc
5.2.5.4. Actual progress - Embedded track
The embedded track standard is the most advanced and has been balloted.
The rewrite consists of the following chapters:
8.1 Introduction (Annex 1)
This section describes the differences between LRV (Light Rail Vehicle) and Streetcars, as they
have different operating environments and thus different track systems (especially curves).
8.2 Track alignment (Annex 2)
Both horizontal and vertical alignment are addressed.
Specific attention was paid to curves and the transitions from tangent to curve and vice versa.
A separate note titled Analysis of Lateral Acceleration and Jerk Rate for Establishing
Superelevation and Spriral Length (Annex 3) provides the background for the curve design.
This note is important as allows and increase of the lateral acceleration from 0.1 g to 0.15 g, which
in turn allows an increase in the allowable unbalanced superelevation (can deficiency) on vurves
and allows correspondingly higher speeds regardless of actual superelevation.
8.4 Rail (Annex 4)
As girder rail or grooved rail is no longer rolled in the US, this section discussion the use of
standard tee rail as an alternative.
It also points out that the girder rails described in the old standard are completely obsolete and
lists and describes to the various new rails that are currently available.
5.2.5.5. Actual progress - non ballasted track
The work on the non ballasted track has only just started and consists of a draft Research Needs
Statement (Appendix A5.2.4).
The document discusses the points of overlap with the embedded track as well as some specific issues:
Stray current protection: track to earth resistance (TTE);
Existing norms and measurement protocols:
o Effects of rebar and type of rebar (coating);
Guarding for curves;
Slab track design and construction:
o Use of rebar or not?
o Plinths and their attachment;
o Top down or bottom up construction;
o Uptimum track stiffness design;
o Transition to and from ballasted track.
TIP5-CT-2006-031312 Page 184 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
APPENDIX A5.2.1: VIBRATION VELOCITY SPECTRA WITH MEASURED
RAIL IRREGULARITIES
VIBRATION SPECTRA FOR THE TRAMWAY VEHICLE
Vehicle in curve (30 m radius) and EB50T rail
Curve at 20 km/h (EB50T)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 185 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in curve (30 m radius) and NP4am rail
Curve at 20 km/h (NP4am)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Vehicle in curve (30 m radius) and 35G rail
Curve at 20 km/h (35G)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 186 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in curve (30 m radius) and Ri59N rail
Curve at 20 km/h (Ri59N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Vehicle in curve (30 m radius) and Ri53N rail
Curve at 20 km/h (Ri53N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 187 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in straight line and EB50T rail
Straight line at 20 km/h (EB50T)
-40
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 40 km/h (EB50T)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 188 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 60 km/h (EB50T)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line on Ballast (EB50T)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 189 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Classic (EB50T)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on CDM-Confort (EB50T)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 190 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Hard (EB50T)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on Direct-Soft (EB50T)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 191 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in straight line and NP4am rail
Straight line at 20 km/h (NP4am)
-40
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 40 km/h (NP4am)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 192 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 60 km/h (NP4am)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line on Ballast (NP4am)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 193 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Classic (NP4am)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on CDM-Confort (NP4am)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 194 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Hard (NP4am)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on Direct-Soft (NP4am)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 195 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in straight line and 35G rail
Straight line at 20 km/h (35G)
-40
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 40 km/h (35G)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 196 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 60 km/h (35G)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line on Ballast (35G)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 197 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Classic (35G)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on CDM-Confort (35G)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 198 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Hard (35G)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on Direct-Soft (35G)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 199 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in straight line and Ri59N rail
Straight line at 20 km/h (Ri59N)
-40
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 40 km/h (Ri59N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 200 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 60 km/h (Ri59N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line on Ballast (Ri59N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 201 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Classic (Ri59N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on CDM-Confort (Ri59N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 202 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Hard (Ri59N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on Direct-Soft (Ri59N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 203 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in straight line and Ri53N rail
Straight line at 20 km/h (Ri53N)
-40
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 40 km/h (Ri53N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 204 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 60 km/h (Ri53N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line on Ballast (Ri53N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 205 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Classic (Ri53N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on CDM-Confort (Ri53N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 206 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Hard (Ri53N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on Direct-Soft (Ri53N)
-40
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
8
0
1
0
0
1
2
5
1
6
0
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 207 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
VIBRATION SPECTRA FOR THE LIGHT RAIL VEHICE
Vehicle in curve (100 m radius) and UIC50 rail
Curve at 20 km/h (UIC50)
-10
-5
0
5
10
15
20
25
30
35
40
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Curve at 40 km/h (UIC50)
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 208 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve at 60 km/h (UIC50)
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Curve on ballast (UIC50)
-30
-20
-10
0
10
20
30
40
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 209 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve on CDM-Classic (UIC50)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Curve on CDM-Confort (UIC50)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 210 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve on Direct-Hard (UIC50)
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Curve on Direct-Soft (UIC50)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 211 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in straight line and UIC50 rail
Straight line at 20 km/h (UIC50)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 40 km/h (UIC50)
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 212 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 60 km/h (UIC50)
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 80 km/h (UIC50)
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 213 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 100 km/h (UIC50)
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line on Ballast (UIC50)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
TIP5-CT-2006-031312 Page 214 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Classic (UIC50)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
Straight line on CDM-Confort (UIC50)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
TIP5-CT-2006-031312 Page 215 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Hard (UIC50)
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
Straight line on Direct-Soft (UIC50)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
TIP5-CT-2006-031312 Page 216 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in curve (100 m radius) and UIC54 rail
Curve at 20 km/h (UIC54)
-30
-20
-10
0
10
20
30
40
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Curve at 40 km/h (UIC54)
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 217 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve at 60 km/h (UIC54)
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Curve on ballast (UIC54)
-30
-20
-10
0
10
20
30
40
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 218 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve on CDM-Classic (UIC54)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Curve on CDM-Confort (UIC54)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 219 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve on Direct-Hard (UIC54)
-30
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Curve on Direct-Soft (UIC54)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 220 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in straight line and UIC54 rail
Straight line at 20 km/h (UIC54)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 40 km/h (UIC54)
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 221 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 60 km/h (UIC54)
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 80 km/h (UIC54)
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 222 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 100 km/h (UIC54)
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line on Ballast (UIC54)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
TIP5-CT-2006-031312 Page 223 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Classic (UIC54)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
Straight line on CDM-Confort (UIC54)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
TIP5-CT-2006-031312 Page 224 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Hard (UIC54)
-20
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
Straight line on Direct-Soft (UIC54)
-30
-20
-10
0
10
20
30
40
50
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
TIP5-CT-2006-031312 Page 225 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
APPENDIX A5.2.2: VIBRATION VELOCITY SPECTRA WITH STEP FUNCTION
RAIL IRREGULARITY
VIBRATION SPECTRA FOR THE TRAMWAY VEHICLE
Vehicle in curve (30 m radius) and EB50T rail
Curve at 20 km/h (EB50T)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 226 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in curve (30 m radius) and NP4am rail
Curve at 20 km/h (NP4am)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Vehicle in curve (30 m radius) and 35G rail
Curve at 20 km/h (35G)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 227 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in curve (30 m radius) and Ri59N rail
Curve at 20 km/h (Ri59N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Vehicle in curve (30 m radius) and Ri53N rail
Curve at 20 km/h (Ri53N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 228 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in straight line and EB50T rail
Straight line at 20 km/h (EB50T)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 40 km/h (EB50T)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 229 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 60 km/h (EB50T)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line on Ballast (EB50T)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 230 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Classic (EB50T)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on CDM-Confort (EB50T)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 231 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Hard (EB50T)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on Direct-Soft (EB50T)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 232 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in straight line and NP4am rail
Straight line at 20 km/h (NP4am)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 40 km/h (NP4am)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 233 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 60 km/h (NP4am)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line on Ballast (NP4am)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 234 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Classic (NP4am)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on CDM-Confort (NP4am)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 235 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Hard (NP4am)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on Direct-Soft (NP4am)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 236 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in straight line and 35G rail
Straight line at 20 km/h (35G)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 40 km/h (35G)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 237 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 60 km/h (35G)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line on Ballast (35G)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 238 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Classic (35G)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on CDM-Confort (35G)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 239 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Hard (35G)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on Direct-Soft (35G)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 240 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Vehicle in straight line and Ri59N rail
Straight line at 20 km/h (Ri59N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 40 km/h (Ri59N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 241 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 60 km/h (Ri59N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line on Ballast (Ri59N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 242 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Confort (Ri59N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on Direct-Hard (Ri59N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 243 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Soft (Ri59N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Vehicle in straight line and Ri53N rail
Straight line at 20 km/h (Ri53N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 244 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 40 km/h (Ri53N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 60 km/h (Ri53N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 245 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Ballast (Ri53N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on CDM-Classic (Ri53N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 246 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Confort (Ri53N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Straight line on Direct-Hard (Ri53N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 247 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Soft (Ri53N)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
VIBRATION SPECTRA FOR THE LIGHT RAIL VEHICE
Vehicle in curve (100 m radius) and UIC50 rail
Curve at 20 km/h (UIC50)
-10
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 248 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve at 40 km/h (UIC50)
-10
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Curve at 60 km/h (UIC50)
-10
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 249 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve on ballast (UIC50)
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Curve on CDM-Classic (UIC50)
0
10
20
30
40
50
60
70
4 5 6 8 10 13 16 20 25 32 40 50 63 79 100 126 158 200
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 250 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve on CDM-Confort (UIC50)
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Curve on Direct-Hard (UIC50)
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 251 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve on Direct-Soft (UIC50)
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Vehicle in straight line and UIC50 rail
Straight line at 20 km/h (UIC50)
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 252 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 40 km/h (UIC50)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 60 km/h (UIC50)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 253 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 80 km/h (UIC50)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 100 km/h (UIC50)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 254 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Ballast (UIC50)
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
Straight line on CDM-Classic (UIC50)
0
10
20
30
40
50
60
70
80
4 5 6 8 10 13 16 20 25 32 40 50 63 79 100 126 158 200
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
TIP5-CT-2006-031312 Page 255 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Confort (UIC50)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
Straight line on Direct-Hard (UIC50)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
TIP5-CT-2006-031312 Page 256 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Soft (UIC50)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
Vehicle in curve (100 m radius) and UIC54 rail
Curve at 20 km/h (UIC54)
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 257 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve at 40 km/h (UIC54)
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Curve at 60 km/h (UIC54)
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 258 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve on ballast (UIC54)
0
10
20
30
40
50
60
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Curve on CDM-Classic (UIC54)
0
10
20
30
40
50
60
70
4 5 6 8 10 13 16 20 25 32 40 50 63 79 100 126 158 200
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 259 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve on CDM-Confort (UIC54)
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Curve on Direct-Hard (UIC54)
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
TIP5-CT-2006-031312 Page 260 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Curve on Direct-Soft (UIC54)
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
Vehicle in straight line and UIC54 rail
Straight line at 20 km/h (UIC54)
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 261 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 40 km/h (UIC54)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 60 km/h (UIC54)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 262 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line at 80 km/h (UIC54)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
Straight line at 100 km/h (UIC54)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
Ballast
CDM-Classic
CDM-Confort
Direct-Hard
Direct-Soft
TIP5-CT-2006-031312 Page 263 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Ballast (UIC54)
0
10
20
30
40
50
60
70
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
Straight line on CDM-Classic (UIC54)
0
10
20
30
40
50
60
70
80
4 5 6 8 10 13 16 20 25 32 40 50 63 79 100 126 158 200
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
TIP5-CT-2006-031312 Page 264 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on CDM-Confort (UIC54)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
Straight line on Direct-Hard (UIC54)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
TIP5-CT-2006-031312 Page 265 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Straight line on Direct-Soft (UIC54)
0
10
20
30
40
50
60
70
80
4568
1
0
1
3
1
6
2
0
2
5
3
2
4
0
5
0
6
3
7
9
1
0
0
1
2
6
1
5
8
2
0
0
Frequency [Hz]
V
e
l
o
c
i
t
y
[
d
B
r
e
f
1
e
-
9
m
/
s
]
20 km/h
40 km/h
60 km/h
80 km/h
100 km/h
TIP5-CT-2006-031312 Page 266 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
APPENDIX A5.2.3 APPLICABILITY OF RAIL STANDARDS
Separate xls-file D0504_D2S_M48_Appendix C.
TIP5-CT-2006-031312 Page 267 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
APPENDIX A5.2.4 EXTRACTS FROM AREMA, CHAPTER 12 PART 8
8.1. INTRODUCTION
This section of Chapter 12 deals specifically with the planning, design, construction and
maintenance of facilities and tracks used for what is commonly called street running, where the tracks
are embedded in pavement or other road surface, and generally the paving surface is even with the top of
rail. Two types of street running track will be covered herein. Type 1 Embedded Track is founded on a
concrete slab, similar to non-ballasted track (covered elsewhere), and the paving infill is usually concrete
or asphalt, but can also be pavers, paving stones, grass, etc.. Type 2 is herein called Paved Track and is
ballasted track of various types (concrete, wood, steel or plastic ties in crushed stone ballast, etc.) covered
with either asphalt, concrete or other type of pavement. Both types of track may be used by a wide
variety of steel wheeled vehicles including light rail vehicles, streetcars, trolleys or trams (the name
depending on local preference), and sometimes shared use with freight trains, and the track structure must
accommodate the types of traffic anticipated, including heavy-axle load rubber-tired traffic.
As the variety of vehicles that might use the tracks covered in this Section of Chapter 12 of the
Manual are myriad, the following verbiage will be used to describe the typical vehicles, viz:
o Light Rail Vehicle (LRV): a vehicle of modern design, sometimes with four axles but
frequently articulated and having six or more axles, used in street running but primarily
intended for relatively high-speed travel between fairly widely spaced stations, often
operated coupled in trains, top operating speed in the 55-65 mph range, and usually limited
to minimum curve radii of 82-83 ft.
o Streetcar: a vehicle of either heritage or modern design, frequently having four axles, but
sometimes articulated and having six or more axles, used primarily in mixed traffic, street
running in downtown circulator operations, based on the tracking capabilities of the
ERPCCar, top operating speed in the 30-45 mph range, and usually capable of negotiating
curve radii down to 35-39 ft.
As all recommendations in this Section are related to hypothetical vehicles, not specific ones, it is
absolutely essential that the designer and specifier be fully conversant with the operating and tracking
capabilities of the vehicle(s) that will actually use the tracks, and to verify suitable track geometric and
alignment criteria that will interact and work properly. It is equally essential that the track designer be
constantly aware that there may be characteristics of the shared street civil or architectural design that
may be detrimental to the design of good and safe track alignment, and that any conflicts should be
resolved as early as possible in the planning..
Embedded track requires special planning and design approaches to integrate the rail facilities
into the urban streetscape successfully and to have the rail vehicles interact efficiently and safely with the
rubber-tired traffic in the shared roadway, while maintaining the appropriate balance between the needs of
the rail transit system and other stakeholders in the busy urban environment. This starts with careful
planning to be sure there are no glaring safety issues caused by the track alignment or facilities and that
the rail vehicles will mesh well with the overall traffic plan and signaling. Further, that the installation in
the streets will not significantly degrade the operation of the rail vehicles, such as excessive street surface
drainage crossfall and curves without spirals. The planning should also include considerations of
ancillary facilities such as locations and designs of overhead contact wire system poles, stations, stops,
traction power substations, pedestrian crosswalks, safety zones, etc.
TIP5-CT-2006-031312 Page 268 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
The design involves developing a comprehensive alignment plan and construction details that
cover the unique requirements of street running, generally described as:
o Types of rail traffic; vehicle loadings and geometric requirements, wayside clearances, safety
issues
o Locations and details for special trackwork, with particular attention to inspection and
maintenance, as well as the interfaces and potential hazards associated with placement of
special trackwork in areas shared with motor vehicles and/or pedestrians.
o Traction power, TP wayside facilities, stray current and corrosion control
o Integration of the track into the street design physically, operationally and esthetically
o Special considerations such as bridges, tunnels, viaducts, especially passenger/pedestrian
safety issues
o Track maintenance inspection, access and repair considerations
o Traffic and rail signal integration; vehicle and pedestrian grade crossings, parking lanes and
safety zones
o Station, stops and their amenities, including safe pedestrian access and protection from auto
and rail traffic
o Maintenance and repair management considerations; life-cycle costs
The planning and the design items listed above are covered in detail in the sections following, and
with references to other Chapters of the Manual and other authoritative sources, as appropriate. These
recommendations are based not only on theory but also on documented experience from both successful
and unsuccessful embedded and paved track and facilities projects and rail transit properties operating
extensive embedded and paved track operations. Where criteria or plans are quoting a specific Agencys
standards, it will be noted, and the reader should be aware that such standards tend to be property-specific
and should be thoroughly investigated as to their appropriateness for any other project. We quote below
advice from the American Transit Engineering Association Engineering Manual, Way and Structures
Division W, issue of 1923:
"This specification is intended to cover the construction of electric railway track in paved city
streets. It is obvious that no general specification can be prepared for such work to cover all special
types of track construction, or to meet special conditions. The scope of this specification has therefore
been limited to an expression of the fundamental principle which should be followed out in constructing
track in paved streets."
AREMA Committee 12 believes that was good advice in 1923, and is still good advice at present,
and therefore we are following it as faithfully as possible in our development of this Manual material.
8.2 TRACK ALIGNMENT
8.2.1 GENERAL
Alignments for embedded track in streets are frequently more constrained than for other LRT
track types (ballasted and direct fixation.) Embedded tracks follow streets within traffic lanes and curb
offsets, make tight turns within street intersections and follow pavement cross sections and profiles.
The primary objectives of any track alignment are cost effectiveness, operating efficiency and
passenger safety and comfort. The alignment recommendations in this section include worst case criteria
for application to embedded track alignment. Like all alignments, the absolute maximum/minimum
TIP5-CT-2006-031312 Page 269 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
alignment criteria herein are to be avoided in favor of longer tangents, flatter curves, and longer spirals
wherever possible. Where the costs of street modifications are minor, they should be incorporated if they
will improve the alignment. Extensive use of absolute maximum/minimum values results in slower
operations and higher maintenance costs.
It is recommended that these worst case criteria be combined with more conservative criteria into
a single criteria document for any specific project. Alignment criteria may be found in Chapter 3 of TCRP
Report No. 57, and in both Chapter 5 Part 3, and Part 3.5 of this chapter of the AREMA Manual.
Developing a general criteria that includes worst case allowable criteria will reduce the time consuming
effort required to grant variances from the general criteria that are often needed otherwise. As stated
above, these worst case criteria should be applied only when general criteria will not produce a feasible
design. Even with comprehensive criteria containing desired values, minimum/maximum values, and
absolute minimum/maximum values, field conditions will occur requiring engineering analysis of
alternatives, judgment and compromise to arrive at a safe, efficient solution.
The criteria in Part 8.2 are based on a typical light rail vehicle (LRV). If possible during
preliminary design, the vehicle parameters affecting alignment criteria should be established. For final
design, it is imperative that the vehicle parameters affecting alignment criteria be established and the
project alignment criteria adjusted accordingly. Due to the greater variance amongst contemporary
streetcars and trolley cars (either new or restored vintage cars) compared with contemporary LRVs, the
advice in this paragraph is of even greater importance for these type operations.
Street running embedded track speeds are usually limited to the legal speed of the roadway traffic which
is seldom over 35mph. For embedded track in open running territory where the typical LRT vehicle is
capable of a sustained operating speed of 55mph or higher, more conservative (lower maximum and
higher minimum) values should be considered for alignment criteria although the absolute
maximum/minimum values given here are still applicable.
Combinations of any of maximum grade, maximum unbalanced superelevation, minimum
horizontal curve radius and minimum vertical curve radius should be avoided. (Further guidance on this
issue is to be developed.)
These criteria assume standard gauge track (56.5 inches.) plus or minus small adjustments for
tight gauge and gauge widening.
Many of the criteria stated herein are excerpted from or derived from information in TCRP
Report No. 57. TCRP Report No. 57, Track Design Handbook for LRT is available from the US
Transportation Research Board (TRB).
8.2.2 VEHICLE INTERFACE
These embedded track alignment criteria reflect the operating limitations of typical modern LRT
vehicles. Circulator system vehicles (street car and trolley car are used synonymously herein) are often
capable of tighter radius horizontal and vertical curves than an LRV.
Individual vehicles may be significantly different in one or more operating characteristics than
the typical values given here. This is known to be specifically true for vehicles with trucks having
independently turning wheels.
TIP5-CT-2006-031312 Page 270 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
These criteria, based on typical values for an LRV, may be considered useful for preliminary
design but they should be adjusted as the actual operating characteristics are established. It is imperative
for an efficient final design that the vehicle specification (or consultant or manufacturer) be consulted as
to vehicle limiting operating characteristics. Alignment criteria for final design must be compatible with
the selected vehicle. Vehicle characteristics should be based on worst case of new or deteriorated
condition. For example, minimum clearance under the vehicle which affects allowable crest vertical curve
radius may be reduced for worn or collapsed suspension compared with new conditions.
For an average, modern, bi-directional, coupled, fully loaded, articulated LRV, typical limiting operating
characteristics are:
Maximum vehicle operating speed 55mph
Maximum allowable grade 7%
Minimum horizontal curve radius 82 feet
Minimum vertical curve radius crest: 820ft, sag: 1150ft
Maximum allowable rate of twist 1 inch in 25ft
Maximum vehicle roll angle < 1.5 degrees (stabilized suspension)
Typical truck spacing 22 to 30 ft
For comparison, typical trolley car limitations are:
Maximum vehicle operating speed 35mph
Maximum allowable grade 9%
Minimum horizontal curve radius 35 feet, centerline of track. (Some
trolley minimum radius criteria are
stated as inside rail radius, which
is also sometimes used to designate
turnout radius.)
Minimum vertical curve radius crest: 310 ft; sag: 560 ft.
Maximum allowable rate of twist 1 inch in 12.5 ft
Maximum vehicle roll angle Varies
Typical truck spacing 22 ft
8.2.3 HORIZONTAL ALIGNMENT
Horizontal alignment consists of tangents, circular curves and spirals in various combinations.
8.2.3.1 Tangents
o The desirable minimum tangent between curves should be the truck spacing plus axle
spacing of a truck (overall wheelbase) so that a vehicle will have adjacent trucks exit one
curve before entering another. No portion of the tangent should be superelevated.
o The absolute minimum tangent between curves is zero so long as the resultant geometry does
not exceed the vehicle coupler maximum angle, the speed does not exceed 20 mph and the
adjoining curves are unsuperelevated.
TIP5-CT-2006-031312 Page 271 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
o If adjoining curves are superelevated, they must have spirals or intervening tangent of
sufficient length to meet superelevation runoff requirements.
o The foregoing criteria apply to reverse curves. For curves in the same direction, a smoother
ride results from a compound curve rather than a short tangent between the two curves.
Compound curves should have spirals connecting the different radius portions of the curve.
The spiral shall begin with the radius of one curve and uniformly increase/decrease to the
radius of the other adjacent curve and not be back to back spirals meeting at a common
tangent.
8.2.3.2 Curves
o The desirable minimum curve radius is 1.5 times the absolute minimum radius.
o The absolute minimum radius is that radius at which a coupled vehicle will negotiate the
curve.
o The desirable minimum length of circular curve (in feet) is three times the normal operating
speed (in mph) of the curve. For spiraled curves this is the length of the circular curve plus
one half the sum of the lengths of the spirals.
o For unsuperelevated spiraled curves the minimum circular curve length is zero, ie back-to-
back spirals.
8.2.3.3 Superelevation
Street running track does not often allow for design of actual superelevation (E
a
) based solely on
operating speeds. While actual superelevation is not precluded on street running track, it is likely that the
superelevation will have to accommodate the cross slope of the street as well as the desired
superelevation. Negative superelevation can occur and speed should be adjusted accordingly.
Since street running requires frequent speed reductions and stops to accommodate street traffic,
the maximum E
a
should not exceed 3 inches. Exceptions to this maximum, such as roadway curves with
larger than 3 inch of cross slope in the roadway and where frequent stopping of trains is unlikely offer
opportunities to use higher actual superelevation. On tangents the maximum cross slope should not
exceed one inch. On tangents and curves, the differential between the street cross slope and track cross
slope/superelevation should not be greater than one inch.
8.2.3.4 Allowable Speed on Curves
Based on ride comfort for short trips and assuming well maintained alignment on LRT embedded
track systems, the recommended lateral acceleration of 0.1g (6 inches of unbalance) may be increased in
critical locations to a lateral acceleration of 0.15g (9 inches of unbalance superelevation).
Allowable speed on a curve is:
_____________
V = E / 0.0007 D
Where V = speed in mph
E = total superelevation in inches, the sum of E
r
+ E
u
+ E
a
Where E
r
= equivalent car body roll allowance which for a stabilized vehicle is 1.5 inches
and for unstabilized suspensions is 3 inches
TIP5-CT-2006-031312 Page 272 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
E
u
= design unbalance: up to 4.5 inches recommended with 7.5 inches maximum for
stabilized suspension cars and up to 3 inches recommended with 6 inches maximum
allowed for unstabilized suspensions
E
a
= actual superelevation in inches.
D = degree of curvature (5730/radius in feet)
8.2.3.5 Spirals
Spirals should be used on all mainline (passenger carrying) embedded track curves. For zero
actual superelevation on embedded track curves, spiral length is determined based on the rate at which
lateral acceleration (unbalance) is introduced. The maximum rate of change of lateral acceleration (jerk
rate) is 0.1g/s. The absolute minimum length spiral L
s
is therefore:
L
s
= 0.29 V E
u
Where Ls = length of spiral in feet
V = velocity in miles per hour
E
u
= unbalance from the curve computation in inches
When curves have superelevation in them, the rate of attainment should not exceed a vertical
acceleration rate of change of 0.1g/s. The equivalent formula is:
Ls = 0.29 V E
a
The ability of the vehicle to withstand twist must also be considered when E
a
is used. For a
typical LRV with an allowable rate of twist of 1 inch in 25 ft, the formula is:
Ls = 25E
a
The longest spiral computed using these three formulae determines the actual spiral length to be
used. The more conservative formulae given in Part 3.5 of this Chapter should be used where they do not
cause excessive cost to implement.
There are many different methods for computing spiral parameters. The notations and formulae in
Chapter 5, Part 3 are recommended for spiral layout computations.
Many different philosophies have been used to proportion E
a
and E
u
on curves. See TCRP Report
No. 57 for applicable formulae.
8.2.4 VERTICAL ALIGNMENT
Vertical alignment is comprised of tangential gradients joined together by parabolic vertical
curves.
8.2.4.1 Tangent Grades
o Maximum gradient must be based on vehicle braking and tractive effort. Typically for LRVs this
requires that sustained grades over 2500 ft long not exceed 6% and shorter sustained grade not exceed
7%.
TIP5-CT-2006-031312 Page 273 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
o Minimum tangent length between vertical curves; desired 100ft; minimum is truck spacing plus axle
spacing on a truck (overall wheel base), usually about 40 ft. Absolute minimum is zero.
o Desirable grade at stations is 0% to 0.35% and in the United States may not exceed 2% in order to
comply with ADA provisions.
8.2.4.2 Vertical Curves
Vertical alignment must follow street grades including crown of cross streets unless the streets
will be re-graded and re-crowned as part of the track construction. The critical vertical curve is the
minimum radius the vehicle will accommodate for sag and crest curves. The minimum vertical curve
must allow for clearance of the underside of the vehicle adjusted for wear and collapsed suspensions.
A typical LRVs vertical curve radius limit is usually around 820 ft for crests and 1150 ft for
sags. Using these values, the equivalent minimum curve length (LVC) can be determined from:
LVC = 0.01AR
Where A = algebraic difference (using the percent grade as whole numbers, i.e. 2.0 % = 2, -2.0% = -2
and 0.35% = 0.35) of gradients connected by the curve, and
R = Limiting radius in ft
For example, crossing a street with a 2% crown (1:50 cross slopes) the minimum LVC = 0.01 x (2 minus
-2) x 820 = 32.8 ft. This length LVC would fit a 40 ft wide street.
The minimum crest LVC is
LVC = AV
2
/25
Where V = design speed in mph
The minimum sag LVC is
LVC =AV
2
/45
Using the above sample curve, the speed for the 32.8 ft long crest LVC should not exceed 32.8 = 4V
2
/25.
V = 14.3 mph.
Back to back reverse curves are acceptable as long as the above minimums are met by each
curve.
Drafted by: Arthur Keffler, 4/7/08
Revised 7/30/08, 8/6/08, 9/17/08, 1/12/09, 1/16/09, and 2/12/09
8.4 RAIL
8.4.1 Rail Considerations
This section discusses rail sections and provides information and recommendations for their
application in embedded track. Both tee rails and grooved rails are used in constructing embedded tracks.
Grooved rails have the advantage of a built-in flangeway, but tee rails are equally satisfactory when
properly applied and more economical, and are now used by most systems for rehab and new
construction. In order of frequency of use, the four major types of embedded track and rail usage in
North America are as follows:
TIP5-CT-2006-031312 Page 274 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
a. Tee rail (nearly all 115RE) in concrete with formed flangeways, frequently installed with
rubber/plastic coverings and sometimes rubber/plastic flangeways that are part of the rail
electrical isolation system, also occasionally embedded in elastomeric grout
b. Grooved rails of various heights and weights, frequently installed with rubber/plastic
coverings and sometimes embedded in elastomeric grout; all are usually part of the rail
electrical isolation system
c. 115RE with a flangeway formed using Strap Guard, mountings similar to a. & b.
d. 115RE on the high side and a grooved guard rail on the low side of guarded curves,
mountings similar to a. & b.
8.4.1.1 Rail Selection Criteria
When considering the specifications for a rail section or sections for use in embedded track the
following six most important considerations should be used to evaluate tee and grooved rails sections:
1. Suitability for the application:
a. beam strength
b. head profile to match wheel profiles and have recommended gauge face angle
c. projected wear life of plain and premium rails
d. height of section which impinges on excavation and paving depth details
e. applicability to the project paving and rail mounting details, and providing a suitable, ADA-
compliant flangeway that is architecturally pleasing and maintanable
f. requirement for guarding, either curves or fully guarding all tracks
g. matching prior rail usage on the property
h. adequate cross section area and conductivity for negative return without excessive voltage
drop
2. Cost factors:
a. first cost
b. premium feature first cost
c. projected life-cycle cost
d. projected future cost for repair or extensions
e. Added cost for guarding devices where needed
3. Availability:
a. rolling frequency
b. projected long-term availability
c. multiple sources preferred
d. availability of premium features and long lengths
e. compliance with Buy America provisions, if applicable
4. Metallurgy & maintenance
a. weldability, electric flash-butt and thermite
b. requirements for special treatment of welds such as post-weld hardening
c. ease of compromise welding to rails of different metallurgy
d. running surface hardness achieved by alloying or heat-treating, or both
e. subject to brittle fracture (especially in cold climates)
f. grinding are grinders available to be used for corrugation removal and re-profiling
5. Adaptability to special trackwork
a. availability of matching cast and/or built-up components
b. adaptability to machining and pre-curving
c. section height suitable for use of asymmetric switch tongues/points
TIP5-CT-2006-031312 Page 275 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
d. suitability for laying in plates or DF Fastenings
6. Quality Assurance
a. Availability of industry recognized quality standards & inspection techniques
b. QC requirements that lend themselves to normal field inspection methods
c. Availability of trained inspectors and suitable equipment to verify the QC requirements
8.4.1.2 Use of Tee or Grooved Running Rails
Based on the criteria above, many properties in North America have selected the 115RE tee rail
section for use in embedded track. The selection was based on the following conclusions:
a. Suitable for most applications regarding strength, head profile, wear life, height, etc.
b. Interfaces well with the AAR 1B wheel profile; reasonably well with ATEA-type wheel
profiles
c. Readily available from several producers; Buy America compliant
d. Initial cost; reasonable delivery times
e. Available head-hardened and in long lengths; some mills furnish CWR
f. Easy to weld, both flash-butt and thermite
g. Some matching special trackwork appliances available
h. Adequate current capacity for most operations
i. Dimensions, properties and quality are controlled by AREMA Chapter 4 specifications,
which are well respected and understood in the industry
It is recommended that the designer or specifier give proper consideration to all the factors listed
above, and apply proper weighting of those factors based on project-specific criteria, including the
historical or aesthetic concerns. The 115RE rail section is normally more cost-effective than grooved rail,
and can be used where practical. Alternatively, other tee rail sections can also be used, such as 85 ASCE,
90 ARA-A, or 100 ARA-B, if available, either new or Class I condition relay. However, there are
situations where grooved rails are preferred, and may have attributes that offset some or all of the
additional cost of the rails, such as:
a. The integral flange guard, in the old days called the tram, provides additional protection
against heel-climb derailments, especially on sharp curves and in special trackwork
b. Having the infill paving, especially asphalt, flush with top of rail on each side reduces the
potential for raveling or chipping and spalling of the pavement
c. The relatively small flangeway opening reduces the tripping hazard for pedestrians and
bicycles vs a large, tooled flangeway
d. Grooved rail is much easier to lay in elastomeric grout embedment, as it doesnt need a
separate flangeway formed in the grout
e. Concrete placement/finishing with modified paving machinery is easier with grooved rail
It should be noted that using tee rail in embedded track requires a means to maintain a suitable flangeway
opening in the infill paving, such as:
a. In Portland cement concrete, a blocked-out, troweled or screeded flangeway of appropriate
dimensions and shape can be easily formed in the concrete
b. In less rigid paving infills, such as hot-mix asphalt, pavers, brick, crushed stone, a flangeway
guarding device will be required such as shown in 8.4.1.3 or a rubber or plastic flangeway
former
c. In rails mounted in polyurethane or similar resilient polymers, a flangeway must be formed
in the polymer, by pouring the polymer low on the gauge side, by use of a flangeway
forming blockout, or a flangeway forming device as shown in 8.4.1.3
TIP5-CT-2006-031312 Page 276 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
8.4.1.3 Typical Flangeway Guarding Methods & Appliances
When tee rail is used, a flangeway can be tooled into the concrete; however, this is not always
acceptable. Therefore, other methods of forming the flangeway are shown in the four figures below.
These are only four of many possible methods, some proprietary, which will produce a satisfactory flange
guard. Where curves are to be guarded, a restraining guard rail device must be added to the tee rail.
Flange guard must not be confused with restraining guard rail; some of the designs shown are configured
correctly and are robust enough to act as restraining guard rail, some are not. A careful choice must be
made as to the appropriate design for service as a restraining guard rail and the specifications should
cover in detail the proper mounting method and required hardware.
Note: There are other methods, as noted above, available to provide a flangeway, some
proprietary. See additional details including electrical isolation techniques in Section 8.5. Installation
Methods
TIP5-CT-2006-031312 Page 277 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
8.4.1.4 Discussion of Girder Rail and Grooved Rail Usage
Historic Domestic Girder, Grooved Girder & Grooved Girder-Guard
Rails
Prior to the 1940s, domestic steel mills produced plain girder, grooved girder or grooved girder-
guard rail sections in several heights and weights, as shown in the cut, above. After the 1980s, when
only two sections were available that were originally intended for freight service, there has not been any
grooved girder rail rolled in North America. None of the Sections shown on AREA Plan 1003-40 are
available new. Therefore, if tee rail with or without a flangeway guarding device/method is not a viable
option, either for practical or architectural reasons, it will be necessary for the designers and builders of
embedded tracks to use grooved rails of non-domestic manufacture. Because this situation has lasted a
long time, the usage of European specialty rails purpose-designed for use as grooved running rails,
restraining guard rails and switch points has become fairly common. Information on most of these rails is
in the following section.
8.4.1.5 European Grooved Rails and Special Trackwork Rail Profiles
The grooved rails produced primarily by European rolling mills are not currently covered by
AREMA specifications; they are covered by several European standards organizations, which control
both the design and manufacture. Section 8.4.2 will provide information on the standards organizations
and their respective specifications and recent changes in those organizations responsibilities. This is
furnished as information only, not an AREMA specification. It is the responsibility of the designers and
users to familiarize themselves with the appropriate, current specifications for rails and special rail
sections contemplated for use in North American projects that will be rolled in non-domestic mills, and
with the terminology used. This section will cover the topics listed below, viz:
TIP5-CT-2006-031312 Page 278 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
a. Information on the standards organizations controlling the specifications applicable to
grooved rails and special rail sections produced primarily in non-domestic mills and to which
AREMA specifications do not presently apply, and limited details of those specifications
and/or recommended practices
b. The changes in nomenclature applicable to certain rails and special sections produced to
European or other standards that are frequently used in North America
c. Typical manufacturing specifications, tolerances and testing of the rails noted in a., and b.,
above
d. Drawings and physical characteristics of certain rails and special sections produced to
European or other standards that are frequently used in North America
e. General recommendations for selection of appropriate rails and special sections
f. Special considerations related to handling, welding (both field and shop), laying and de-
stressing of embedded rails
8.4.2 Standards Organizations and Relevant Standards or Recommended Practices
8.4.2.1 Standards Organizations
a. UIC - The International Union of Railways (UIC is a French acronym for, LUnion
Internationale des Chemins de fer) is an international organization based in France whose
purpose is to promote the interests of railway transport on a worldwide basis, including
technical cooperation. Prior to the creation of the European Union, many rail standards were
controlled by the UIC, and rail sections were named with UIC in the nomenclature, such as
UIC-60 and UIC-33. That role is now filled by the CEN (see below). The UIC is similar to
the AAR combined with the focus on passenger transport of APTA.
b. VDV The Association of German Transport Undertakings (in German = Verband
Deutscher Verkehrsunternehmen; formerly VV) is an organization of German-speaking
public transit and freight rail groups to provide cooperative technical guidance similar to
AREMA; the specifications they publish are recommended practices, not standards.
Grooved rails were, and in some cases still are, supplied per VDV specifications, and
tramway special trackwork is still controlled by VDV.
c. CEN The European Committee for Standardization (languages: English, German, French)
is based in Brussels, Belgium, and publishes standards for a multiplicity of technical
endeavors, including rails controlled by the steel committee. The CEN is like ASTM, ACI,
ASME, IEEE, SAE, AAR, AREMA, etc. rolled into one standards organization. The
signatory countries, now more than thirty, are required to accept the European Norm
standards as their own without alteration. These standards have EN plus an identification
number and date of approval in the name; in the case of grooved rails and special
construction sections, the CEN standard is EN 14811:2006, which replaced both UIC and
VDV specifications in most cases. If the standard has pr before the name, such as prEN
14811:2006, that indicates a provisional status; the provisional standard has been approved
by the sponsoring committee, but has not been approved by all the signatory countries.
However, it is generally considered to be in effect as approved standards drafted by the
designated controlling committee(s) are seldom rejected by the signatories. The information
following is based primarily on the CEN EN 14811:2006 standard with some additional
information from CEN standard EN 13674:2005 which covers tee (Vignole, also called flat
bottom) rails and special sections of interest such as restraining guard rails, STW
construction rails, and asymmetric switch point sections.
And for domestically produced tee rails the relevant standards are controlled by:
a. American Railway Engineering and Maintenance-of-Way Association (AREMA) domestic
tee rails only
TIP5-CT-2006-031312 Page 279 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
b. American Society of Civil Engineers (ASCE) lighter tee rail sections, mainly 85 AS,
primarily an industrial section, but rolled regularly
8.4.2.2 Applicable European CEN Standards EN 14811, EN 13674, and VDV
The nomenclature of grooved rails and certain special construction rail sections have been
standardized and harmonized per Table 1, below. All drawings, plans, specifications and procurement
documents should reflect the proper CEN Standard nomenclature, where applicable, to avoid confusion
and errors. If the profiles are per VDV standards, the same information noted should appear in all
documents.
Table 1
Revised Standard Nomenclature of Grooved Rails and Construction
Rails per CEN Standards EN-14811 and EN-13674 (Partial List)
CEN Standard
Profile Designation
Prior Profile Designations
(VDV, UIC, etc.)
Generally Applicable To Fig.
51R1 Ri 52-R13, Ri 52 Running rails, H = 130 mm 1
53R1 R1 53-R13, Ri 53 Running rails, H = 130 mm 2
55G1 35 GP Running rails, H = 152.5 mm 3
56R1 Ri Ic Running guard rails, H = 160 mm 4
59R1 Ri 59-R10, Ri 59 Running rails, small g.c. radius, H = 180 mm na
59R2 Ri 59-R13, Ri 59N Running rails, large g.c. radius, H = 180 mm 5
60R1 Ri 60-R10, Ri 60 Running rails, small g.c. radius, H = 180 mm na
60R2 Ri 60-R13, Ri 60 N Running rails, large g.c. radius, H = 180 mm 6
62R1 NP4aMod Running guard rails, H = 180 mm 7
67R1 Ph 37a Running rails, large flangeway, H = 180 mm 8
49E1A1 Zu2-49 Switch tongue profile, H = 116 mm 9
61C1 Ri Ii STW const. grooved rail, flange-bearing, H = 160 mm 10
75C1 BA 75 STW const. grooved stock rail, H = 180 mm 11
76C1 VK Ri 60 STW const. blind groove guard rail, H = 180 mm 12
33C1 U69, UIC33, RI 1-60 Frog guard & restraining rails, H = 93 mm 13
Fz 36 Fz 36, Zu 36 Switch tongue profile, H = 75 mm 14
GGR-118 GGR-118 Running grooved guard rails, H = 168,3mm (6.625-in) na
Footnotes:
1) Section is not controlled by standards; produced per producing mills and/or users designs & specs
2) Section is no longer rolled but is in track on several NA properties, as info only; not a CEN standard
3) g.c. is gauge corner
General notes:
1. The rail sections listed above either are being or have been used in North America with some
regularity.
2. Not all sections listed in Table 1 are illustrated on the following pages. In addition, many more
rail sections (profiles) not listed here are available from some manufacturers. Those have not
been included here because they have either not been adopted as CEN standards or have seen
TIP5-CT-2006-031312 Page 280 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
little or no use in North America. For other sections available, refer to mill catalogues and to the
referenced CEN standards, or other standards if not covered by CEN.
3. Sections including the letter R in their designation are grooved rails; in German,
Rillenschiene. Grooved rails, commonly known in North America as girder rails, are rolled
with an integral flangeway in the head of the rail, and used in the construction of ordinary
embedded track. Sections including the letter C in their designation are known as
construction rails and used in the fabrication of special trackwork. When fabricating STW
using construction rails, the flangeways and head contours are machined (see Fig. 12).
Additional details of embedded special trackwork construction are in 8.7, presently under
development, which also covers designs outside the scope of CEN and VDV standards.
4. For a more detailed discussion of the application of grooved rails in LRT construction,
especially as relates to wheel profile/rail groove matching, please see Transit Cooperative
Research Program (TCRP) Report #57 Light Rail Track Design Handbook.
8.4.2.3 Rail Profile Drawings with Properties of the Sections/Profiles
Figures 1 through 14 shown below have the principal dimensions called out, along with basic
section properties. Some sections are not shown where they are almost identical to another section, with
the key differences noted. For complete dimensions and properties, please refer to the appropriate CEN
or VDV Standard, or the producing mills drawings or catalog.
Fig 1 is typical of the drawing style & dimensioning to be used for all rail sections shown; these
figures are provisional only until artwork is completed
TIP5-CT-2006-031312 Page 281 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
Note: Profiles 59R1 and 60R1 are similar to 59R2 and 60R2, respectively, except that the gauge
corner radius is 10mm (0.394-in), rather than 13mm, and the flangeway is approximately 3-5mm
narrower
TIP5-CT-2006-031312 Page 282 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
TIP5-CT-2006-031312 Page 283 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
TIP5-CT-2006-031312 Page 284 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
8.4.2.4 Properties of Grooved and Construction Rail Profiles/Sections
Table 2
Table of Properties - Grooved Rails and Construction Rails per CEN Standards EN-14811 and EN-
13674, & Some Proprietary Items (Partial List)
CEN
Profile kg/m lb/yd cm in cm
4
in
4
cm
4
in
4
cm in cm in m/tonne ft/gr ton
51R1 51.37 103.6 65.44 10.14 1289 30.93 695.6 16.69 198.4 12.10 198.1 12.08 19.47 64.86
53R1 52.98 106.8 67.49 10.46 1326 31.82 738.4 17.72 208.3 12.71 199.9 12.19 18.87 62.92
55G1 54.77 110.4 69.78 10.82 2076 49.82 681.5 16.36 285 17.39 260.5 15.89 18.26 60.87
56R1 55.98 112.9 71.31 11.05 2477 59.45 802 19.25 349 21.29 278 16.96 17.86 59.52
59R1 58.97 118.9 75.12 11.64 3267 78.41 886.2 21.27 373.8 22.80 352.8 21.52 16.96 56.52
59R2 58.14 117.2 74.07 11.48 3211 77.06 757 18.17 363.1 22.15 350.5 21.38 17.2 57.32
60R1 60.59 122.2 77.19 11.96 3353 80.47 928.6 22.29 391.4 23.88 355.4 21.68 16.5 54.99
60R2 59.75 120.5 76.11 11.80 3298 79.15 920.1 22.08 380.6 23.22 353.3 21.55 16.73 55.76
62R1 62.37 125.8 79.45 12.31 3535 84.84 1042 25.01 427.6 26.08 363.3 22.16 16.03 53.42
67R1 66.76 134.6 85.04 13.18 3554 85.30 1250 30.00 436 26.60 360.8 22.01 14.98 49.92
49E1A
1
63.14 127.3 80.43 12.47 1098 26.35 681.9 16.37 165.3 10.08 221.7 13.52 15.83 52.76
61C1 60.79 122.6 77.44 12.00 2631 63.14 834 20.02 394 24.03 283 17.26 16.45 54.82
75C1 75.23 151.7 95.84 14.86 3596 86.30 967.5 23.22 398.3 24.30 400.8 24.45 13.29 44.29
76C1 72.73 146.6 92.65 14.36 3949 94.78 1049 25.18 529.6 32.31 374.6 22.85 13.75 45.82
33C1 32.99 66.52 42.02 6.51 297 7.13 218.8 5.25 83.7 5.11 51.8 3.16 30.31 101.01
Fz 36 33.99 68.53 46.8 7.25 933.7 22.41 1190 28.56 NA #### NA #### 29.42 98.05
GGR-118 58.3 117.6 74.3 11.52 2640 63.43 777 18.65 321.5 19.6 283.2 17.3 17.15 57.16
S base
Rail Lg/Unit Wt Linear Mass Area
I xx (Note 2) I yy S head
Note 1: values are valid to only three significant figures at this writing; they should be verified prior to
performing stress calculations and writing firm procurement or construction specifications
Note 2: Some sections show the Moment of Inertia to the I
X-X
axis (the base), not the I
x
Neutral Axis; see
appropriate producers drawing to verify the geometric properties of the section/profile of interest
8.4.2.5 Manufacturing Methods, Tolerances and Testing
All European specifications for grooved rails and construction rails require the use of steel
produced by the continuous casting process, with vacuum-degassed steel specified for rails to be head-
hardened; however, there are some substantial differences in the philosophy behind the specifications, viz
1. Manufacture:
a. EN 14811 is performance-based, rather than prescriptive, wherever possible
b. The six grades of non-alloyed rail steels are classified by hardness, not tensile strength; three
grades are as-rolled, three grades are heat-treated
c. The hardnesses specified range from 200-240 HBN to 340-390 HB
d. The allowable mass of included hydrogen is specified for each grade in PPM, and is controlled
by testing the blooms
e. Alloyed rails are covered by agreement between customer and producer
f. EN 14811 references other CEN standards to specify steel grade nomenclature, and tensile and
hardness testing
g. Quality management is based on the producer adhering to the requirements of EN ISO 9001
TIP5-CT-2006-031312 Page 285 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
2. Tolerances:
a. Rails are produced to two different tolerance levels, analogous to railroad vs industrial quality
b. Many more measurement points on the profile are required in EN 14811 than prior standards
c. The profile and straightness tolerances are generally greater in EN 14811 than in AREMA
Chapter 4, Table 4-2-2 (ie: in EN 14811, height of rail 0.059-in [1,5mm] vs Chap. 4 + 0.030-
in. [0,76mm]/- 0.015-in [0,38mm] based on the premise that the traffic is relatively low speed
d. Construction rails used in making special trackwork have tighter tolerances than running rails
e. Both minor upsweep and downsweep are acceptable
f. Rail length tolerance is much tighter than Chap. 4
3. Testing:
a. Testing procedures are generally similar to AREMA practice
b. For the as-rolled profiles, hardness testing is required on the running surface only; for heat-
treated, both running surface and internal hardness testing is required
c. Purpose-designed gauges are used for profile checking
d. No tests are specified to determine residual stresses
8.4.2.6 Additional Considerations for Grooved Rail Selection
a. The selection criteria listed in Section 8.4.1.1 are equally applicable to grooved rails of non-
domestic manufacture,
b. Investigate the popularity of a candidate profile/section regarding how often it is rolled, by how
many producers, etc., as this has important implications regarding long-term availability and
cost
c. Determine the chemical composition and hardness of a candidate section to make sure that
welding will not be difficult or require special procedures, such as post-hardening; if special
procedures are required, make sure they are covered in the construction specifications
d. Obtain proper handling information from the producer regarding slinging long rails with
spreaders and put this information in the specs
e. Determine compatibility of candidate section with the wheel profile(s) to be used; note, for
instance, that 59R1 and 59R2 have different gauge corner radii and slightly different groove
widths, important considerations in sharp curves
f. Determine that shipment of a candidate section will be done so as to protect the rails from salt-
spray corrosion during transit, and that an appropriate spreader is available to unload the rails
without damage
8.4.2.7 Special Considerations Regarding Handling, Welding, Laying and De-stressing Rails
a. The recommendations in Chapters 4 and 5 should be followed faithfully, plus some special
considerations listed below
b. Handling: special care should be taken when lifting or moving grooved rails, as the thin web
and base flanges make it easy to cripple the base or web if the rails are overbent in handling
(see 8.4.2.6.c and 8.4.2.7.d), or to twist it beyond the yield point
c. Welding: care should be exercised in both flash-butt and thermite welding to make sure the web
and base are not overheated, or base droop and/or web curling may occur
d. Laying: welded strings should not be dragged around sharp corners or otherwise mishandled as
noted in 8.4.2.7.b to prevent kinking or twisting the rails
e. De-stressing: there is no common agreement at this time whether embedded rails need de-
stressing in the conventional sense specified for open track, as sun-kinks are not likely;
however, it is prudent to lay the rails at something near the average ambient temperature to
reduce any tendency to have pull-aparts. This practice is also recommended for all running rails
in embedded tracks.
TIP5-CT-2006-031312 Page 286 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
APPENDIX A
Commentary on Analysis of Lateral Acceleration and Jerk Rate for Establishing Superelevation
and Spiral Length
Introduction
The US rail industry standard for lateral acceleration and jerk respectively for a long time has
been 0.1g and 0.03g/s. The standard used by railroads and transit properties in the US is based on research
conducted 50 years ago and was applicable to all types of cars including dining cars where a smooth ride
was essential. Today, several European countries allow higher rates. SNCF (French National Railways)
uses 0.15g for lateral acceleration and 0.1g/s for jerk for its railroads including the TGV system. Some
higher values for jerk rate have been suggested by research on high speed rides but do not seem to have
been put into practice. Subjective experiments of ride comfort on curves were judged as "noticeable
lateral acceleration" at 0.1g and "strongly noticeable but not uncomfortable" at 0.15g. For short LRT
rides, strongly noticeable lateral acceleration now and then would seem to be an acceptable ride
condition. While the data is less conclusive for jerk, several studies support a higher rate with some
research suggesting it is not a factor in ride comfort at all. It therefore seems reasonable to consider a
somewhat higher jerk rate as well.
Increasing maximum allowable lateral acceleration equals use of a higher limit for unbalanced
superelevation (cant deficiency) on curves and correspondingly higher speeds regardless of actual
superelevation.
Jerk rate is one of three parameters (jerk, twist, and rate of twist) used to establish minimum
spiral length. Allowing a maximum higher jerk rate will allow shorter spirals. In unsuperelevated curves
common to embedded track, jerk is the only parameter used to determine spiral length.
Various researchers from Hirshfeld (1932) and Code (1955) to more recent studies for high speed
rail travel in the US, France, Germany and Japan (1989 to 2004) have examined ride comfort versus
unbalanced superelevation on curves and jerk rates for spirals. The results of those studies produced
recommended rates that range from less than 0.1 g to 0.16g for lateral acceleration. For jerk rate, the
studies recommendations range from 0.03 g/s to 0.25g/s with additional other limitations for the higher
jerk rates. Analyses of ride comfort relative automobile and airplane performance under situations
somewhat analogous to railroad curving have been made. Analyses of ride comfort versus vibration levels
and uneven ride conditions (lateral jolt due to track irregularities) have also been made and comparisons
made to ride comfort on railroad curves. The overall conclusion of these studies is that severe jolts and
long term vibrations have more to do with rider comfort than reasonable lateral acceleration levels and
spiral jerk rates. Safety (rather than comfort) limits were examined in one report which suggest that as jolt
rates (and spiral jerk rate) increase, the lateral acceleration must be decreased so that the two in
combination do not produce an unsafe ride. Unsafe meaning some standing riders would loose their
footing.
Ride comfort is a subjective parameter, and while for the sake of analysis, it is equated with
precise values of acceleration (g) it is not really a precise parameter. Ride comfort is affected by the
vehicle characteristics as well as the track design. Vehicle characteristics vary significantly from one
design to another. Code used a wide variety of passenger cars in his ride comfort studies and in the end,
simplified the varying performance of the cars into just two classes, those with loose suspensions and
TIP5-CT-2006-031312 Page 287 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
those with stabilized suspensions. These two factors the subjective nature of ride comfort evaluation
and the variability of the cars to affect ride comfort - make research to establish values for all systems
problematic. A better approach is to evaluate ride comfort for a given system by operating its vehicles at
varying speeds around a number of curves to decide, for the specific system what constitutes a
comfortable ride.
Lateral Acceleration Discussion
Ride comfort on the body of a curve is determined from a combination of vehicle roll and
unbalance of the curve. A 0.1g value is equivalent to 6 inches total unbalance. For a loosely sprung
vehicle, up to 3 of those 6 inches is consumed by vehicle roll leaving 3 inches E
u
as a maximum design
value for alignment criteria. For more stable cars (ie those with suspensions that limit roll to 1.5
degrees or less per AREMA Chapter 5 test procedure), the E
u
max for design rises to 4.5 inches since the
vehicle roll uses 1.5 inches or less of the combined total of 6 inches allowable unbalance.
SNCF uses 0.15g for lateral acceleration. This has been a suggested acceptable level by others in
the US but does not appear to have been implemented elsewhere. SNCF also commits to maintaining
track alignment to limit lateral jolts due to misalignment to less than 0.025g/s. An FRA sponsored 1991
ride safety (not comfort) study indicates that it is safe to operate at speeds equal to 0.15g lateral
acceleration if track alignment is well maintained so as not to introduce excessive jolts due to
misalignment into the ride. It concluded up to 0.183 g/s jolt with 0.15g lateral acceleration as safe. The
safety study was based on analysis of ride quality on many curves at various speeds. Recent TTCI (2008)
research demonstrated that wheel climb derailment potential is virtually unaffected by unbalance whereas
lower rail angle and track perturbations are the principle causes of wheel climb.
Using the higher 0.15g value for lateral acceleration allows increasing the total unbalance from 6
to 9 inches. This allows a 3inch E
u
design limit for loosely sprung vehicles to rise to 6 inches E
u
and a 4.5
inch E
u
design limit for stabilized suspension vehicles to rise to 7.5 inches E
u
. For an 82 ft radius
unsuperelevated curve (typical embedded track street corner turn and loop situations) the 3 to 6 inch E
u
increase, when using the standard formula E = 0.0007V
2
D to compute velocity (V), improves the speed
from 8 to 11 mph (approx) and a 4.5 to 7.5 inch E
u
increase improves the speed from 10 to 12mph.
Use of E
u
for the E value on unsuperelevated curves without adding the unbalance component
due to car roll results in a conservative (lower) speed and lower g value being computed when using the
standard formula E = 0.0007V
2
D to compute velocity (V). See note at end of this section on D vs R. This
is regardless of combination of E
a
plus E
u
generally considered to be the total E value. The design formula
should include an E
r
(for roll) value when computing V. Since less than the full E value (E
u
+ E
a
+ E
r
) is
used in the current commonly used design formula, the lateral acceleration for the design speed is below
0.1g. Another way to look at this is that it is safe to increase the running speed of a curve without
exceeding the ride comfort for which it was designed although the faster one goes the more
uncomfortable the ride becomes. By adding the E
r
value, the design formula becomes E
a
+E
u
+E
r
=
0.0007V
2
D. The inclusion of 1.5 inches for a stabilized vehicle for E
r
in the formula further improves the
design allowable speed for an 82ft unsuperelevated curve to 14 mph from 12 mph.
The matter of whether or not increasing allowable lateral acceleration increases the risk of wheel
climb derailment or overturning should be considered. Even with shallow wheel flanges, wheel climb is
caused by wheel/rail angle, angle of attack, and suspension stiffness. Lateral acceleration due to
TIP5-CT-2006-031312 Page 288 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
speed (unbalance) if increased indefinitely leads to vehicle overturning not wheel climb. This is so
because the higher lateral force on the wheel due to higher lateral acceleration is offset by more of the
vehicle weight transferring to the vertical component on the wheel. TCRP Report No. 57 has formulae for
analyzing overturning which may be used for comparison with the proposed lateral acceleration/Eu
values. Using the TCRP formula, the overturning speed for an 82ft radius, unsuperelevated curve is 26
mph or about twice the proposed operating speed of 14mph. The safe speed is defined in TCRP Report
No. 57 as the speed at which the vehicle becomes unstable and in danger of derailment upon introduction
of any anomaly in the track. The maximum safe Eu value, using the TCRP formula for safe speed is 9.6
inches Eu which is greater than the proposed maximum of 9 inches. Embedded track, once built to
accurate alignment, should retain the accuracy of that alignment indefinitely.
In summary, increasing the allowable lateral acceleration from 0.1g to 0.15g and adding 1.5
inches for well suspended vehicles for the E
r
value to the formula for computing speed will result in an
allowable safe increase in speed from 10 to 14 mph for an 82 ft radius unsuperelevated curve. Based on
observations by trackwork engineers riding on various LRT systems, this modest adjustment to the design
criteria will do no more than reflect actual operating conditions on systems where operators frequently
increase speed before a train has cleared a curve.
(Note: The standard formula E = 0.0007 V
2
D uses D based on D = 5730/R. This formula was derived
when curves were surveyed with transit and tape methods and defining a curve by Degree of Curve was
useful in the field for staking curves. As noted in surveying texts this method of staking a curve becomes
progressively more inaccurate as radius of curve decreases. It is accurate, however for converting R to D
for use in the above formula for computing speed (V) or total superelevation (E) even at the small radii
anticipated for LRVs and trolley cars. In other words, D should not be used to define the radius of a
curve of less than 300 feet but may be used to convert R to D in the above formula.)
Jerk Rate Discussion
The jerk rate establishes the time needed to introduce the lateral acceleration or unbalance of a
curve at the beginning and end of a circular curve. A constantly increasing amount of lateral acceleration
beginning at zero and ending at the desired lateral acceleration value for a curve is achieved through the
passage, at a constant speed, of a vehicle traveling along a constantly increasing degree of curvature, ie a
spiral.
The length of the spiral determines the time required to go from zero lateral acceleration to the
level of lateral acceleration of the circular curve. It has been demonstrated that the amount of lateral
acceleration (E
u
) is more important to ride comfort than the rate at which it is introduced (spiral length).
Never the less, an unreasonably high rate of introduction of lateral acceleration (jerk rate) is undesirable,
especially for high levels of E
u
. If no spirals are used, the jerk rate is theoretically infinite. In reality, the
play between the wheels and track gage along with dynamic response of the vehicle reduces this infinite
rate to a jerk rate that is measurable though high.
The current US standard of 0.1g lateral acceleration, coupled with an 0.03g/s jerk rate dictates the
introduction of E
u
over 3.33 seconds. This, for a typical maximum E
u
of 4.5 inches transforms into the
familiar formula for determining spiral length: L
s
= 1.09VE
u
.
TIP5-CT-2006-031312 Page 289 of 289
URBAN TRACK Issued: May 28, 2010
D0602_M48_UITP_SP5.doc
In the US, just as with lateral acceleration, a conservative low jerk rate of 0.03 g/s was adopted as
standard. However, numerous studies, beginning with Hirshfeld, concluded that higher jerk rates were
acceptable with respect to ride comfort. The FRA Ride Safety Study of 1989 concluded that jerk rate was
not significant to ride comfort and that rates (either jolt or jerk) as high as 0.183 g/s were safe for lateral
acceleration values up to 0.15g. The 1978 NEC study of ride comfort endorsed the SNCFs values of
0.15g and 0.10 g/s with a limit on jolt of 0.025 g/s. The 2004 FRA Study for high speed rail between
Richmond and Charlotte endorsed the same SNCF values and noted that a jerk rate as high as 0.25 g/s
would be acceptable so long as no track irregularities were to occur that would momentarily raise the jerk
rate to a higher level.
The conclusion from these studies is that a jerk rate of 0.10 g/s would not produce an
unacceptable ride on embedded track which, once properly constructed to a smooth alignment, would
preclude any unusual jolt values from occurring. In fact, a jerk rate of 0.10 g/s is conservative compared
with some recommendations.
For a lateral acceleration maximum of 0.15g, a jerk rate of 0.1g/s means the spirals will be long
enough to introduce the E
u
over 1.5 seconds. The spiral formula for a 0.15 g lateral acceleration (7.5
inches E
u
) and 0.10 g/s jerk rate becomes L
s
= 0.29VE
u
.
To put this in perspective an 82 ft radius unsuperelevated curve used by a stabilized suspension
vehicle could have the following designs:
a. Existing standard of 0.1g and 0.03g/s: E = 4.5 inches with roll ignored.
E = 0.0007 V
2
D V = 9.6 mph L
s
= 1.09VE
u
L
s
= 47.1 ft
b. Proposed rates of 0.15g and 0.10g/s Total E, including roll, of 9 inches, design E
u
of 7.5 inches.
E = 0.0007 V2 D V = 13.6 mph Ls = 0.29VEu Ls = 29.6 ft.
It should also be noted that the lateral offset of the unspiraled versus spiraled curve is 1.12 ft for a.
and 0.46 ft for b.
Conclusions
Based on the foregoing analysis of research conducted by others plus observations of actual
operating conditions, the following is recommended.
1. That the allowable lateral acceleration of 0.1g be increased to 0.15g which corresponds to an
increase of 3 inches of allowable unbalance for a total E
u
of 9 inches (7.5 inches for design plus
1.5 inches for roll).
2. That the E value used in the formula E = 0.0007 V
2
D include the allowance for vehicle roll,
which for modern LRVs would be 1.5 inches along with the new design value for E
u
of 7.5
inches. These first two recommendations will result in higher allowable speeds on curves.
3. That the allowable jerk rate for spiral design be increased from 0.03 g/s to 0.10 g/s. The formula
for computing spirals will be L
s
= 0.29 V E
u
for a maximum of 7.5 inches E
u
. This will result in
shorter spirals with correspondingly smaller offsets.
4. Application of these higher values for lateral acceleration and jerk imply a commitment to high
quality construction and maintenance of track alignment. They should be used only where
allowed by field conditions.
5. These recommendations apply only to standard gauge embedded track.