SIST EN 17936:2024

SLOVENSKI STANDARD
01-december-2024
Železniške naprave - Akustika - Merjenje osnovnih pogojev za izračun okoljskega
hrupa
Railway applications - Acoustics - Measurement of source terms for environmental noise
calculations
Bahnanwendungen - Akustik - Messung der Quellterme für
Umgebungslärmberechnungen
Applications ferroviaires - Acoustique - Mesurage des termes sources pour le calcul du
bruit en environnement
Ta slovenski standard je istoveten z: EN 17936:2024
ICS:
17.140.30 Emisija hrupa transportnih Noise emitted by means of
sredstev transport
93.100 Gradnja železnic Construction of railways
2003-01.Slovenski inštitut za standardizacijo. Razmnoževanje celote ali delov tega standarda ni dovoljeno.

EN 17936
EUROPEAN STANDARD
NORME EUROPÉENNE
October 2024
EUROPÄISCHE NORM
ICS 17.140.30; 93.100
English Version
Railway applications - Acoustics - Measurement of source
terms for environmental noise calculations
Applications ferroviaires - Acoustique - Mesurage des Bahnanwendungen - Akustik - Messung der Quellterme
termes sources pour le calcul du bruit en für Umgebungslärmberechnungen
environnement
This European Standard was approved by CEN on 2 September 2024.

CEN members are bound to comply with the CEN/CENELEC Internal Regulations which stipulate the conditions for giving this
European Standard the status of a national standard without any alteration. Up-to-date lists and bibliographical references
concerning such national standards may be obtained on application to the CEN-CENELEC Management Centre or to any CEN
member.
This European Standard exists in three official versions (English, French, German). A version in any other language made by
translation under the responsibility of a CEN member into its own language and notified to the CEN-CENELEC Management
Centre has the same status as the official versions.

CEN members are the national standards bodies of Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia,
Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway,
Poland, Portugal, Republic of North Macedonia, Romania, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Türkiye and
United Kingdom.
EUROPEAN COMMITTEE FOR STANDARDIZATION
COMITÉ EUROPÉEN DE NORMALISATION

EUROPÄISCHES KOMITEE FÜR NORMUNG

CEN-CENELEC Management Centre: Rue de la Science 23, B-1040 Brussels
© 2024 CEN All rights of exploitation in any form and by any means reserved Ref. No. EN 17936:2024 E
worldwide for CEN national Members.

Contents Page
European foreword . 4
Introduction . 5
1 Scope . 6
2 Normative references . 6
3 Terms and definitions . 7
4 Instrumentation and calibration . 10
4.1 General. 10
4.2 Acoustic instrumentation . 10
4.3 Calibration of the acoustic instrumentation . 11
4.3.1 Temporary setup . 11
4.3.2 Automated calibration . 11
4.4 Non-acoustic instrumentation . 11
4.4.1 Time and duration . 11
4.4.2 Speed, track and direction . 12
4.4.3 Meteorological parameters . 12
4.4.4 Combined roughness . 12
4.4.5 Wheel and track parameters . 12
5 Approach to derive source terms . 12
5.1 General. 12
5.2 Procedure . 13
5.3 Directivity . 14
5.4 Distribution over source heights . 14
6 Measurement procedures . 14
6.1 General. 14
6.2 Rolling noise . 17
6.2.1 General. 17
6.2.2 Measurement steps . 17
6.2.3 Network and fleet roughness . 18
6.3 Traction and equipment noise . 18
6.3.1 General. 18
6.3.2 Source level for traction/equipment noise following EN ISO 3095:2013 . 19
6.3.3 Source level for traction/equipment noise using statistical data collection . 19
6.3.4 Source separation . 19
6.4 Impact noise . 20
6.5 Curve squeal . 20
6.6 Bridge noise . 21
6.7 Braking noise. 22
6.7.1 General. 22
6.7.2 Source level for braking noise following EN ISO 3095:2013 . 22
6.7.3 Source level for braking noise at arbitrary sites . 22
6.7.4 Source level for braking noise at higher speed . 23
6.8 Aerodynamic noise . 23
7 Sampling requirements . 23
7.1 Practical validity . 23
7.2 Site requirements and selection . 24
7.3 Train selection . 24
7.4 Train speeds . 25
7.5 Numbers of pass-bys . 25
7.6 Frequency range . 25
8 Processing . 25
9 Uncertainties . 32
10 Reporting . 32
Annex A (informative) Methods to determine combined effective roughness . 34
A.1 General . 34
A.2 Direct method requiring track and vehicle possession . 34
A.3 Indirect method using rail vibration – trackside method . 34
A.4 Survey methods . 34
Annex B (informative) Calculation of sound power from sound pressure . 36
B.1 Definition . 36
B.2 Calculation method . 36
B.3 Tabulated transfer functions for CNOSSOS for given track geometries and source-
receiver combinations . 37
Annex C (informative) Separation methods for rolling noise of vehicle and track . 62
C.1 Principle of separation . 62
C.2 Separation based on calculation . 62
C.3 Separation based on measurement . 63
C.4 Separation based on external reference . 63
Annex D (informative) Specific environments . 64
Annex E (informative) Transposition between train and track types . 65
Bibliography . 66
European foreword
This document (EN 17936:2024) has been prepared by Technical Committee CEN/TC 256 “Railway
applications”, the secretariat of which is held by DIN.
This European Standard shall be given the status of a national standard, either by publication of an
identical text or by endorsement, at the latest by April 2025, and conflicting national standards shall be
withdrawn at the latest by April 2025.
Attention is drawn to the possibility that some of the elements of this document may be the subject of
patent rights. CEN shall not be held responsible for identifying any or all such patent rights.
Any feedback and questions on this document should be directed to the users’ national standards body.
A complete listing of these bodies can be found on the CEN website.
According to the CEN-CENELEC Internal Regulations, the national standards organisations of the
following countries are bound to implement this European Standard: Austria, Belgium, Bulgaria, Croatia,
Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland,
Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal, Republic of North
Macedonia, Romania, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Türkiye and the United
Kingdom.
Introduction
In Europe, various prediction methods for environmental noise exist. For noise mapping and planning of
railway lines, the prediction methods are enshrined in national or European legislation such as Directive
(EU) 2015/996 (CNOSSOS). An integral aspect to ensure realistic results is to use valid input data.
Environmental noise prediction models consist of noise source terms and a propagation model.
The calculation of railway traffic sources is based on traffic data such as train types, speeds and flow, and
on the vehicle/track noise source terms.
In this document, measurement methods are specified for the acoustic input parameters for the
vehicle/track noise source terms. The collection of traffic flow data are outside the scope of this
document. The method can be used to collect data from different railway noise source types, within the
practical constraints of widely available measurement equipment and railways in normal service.
The document covers the measurement of separate physical source types which are listed in the scope.
Each source type is characterized in terms of its frequency spectrum (up to one-third octave band
details), source height and directivity.
The description of rolling noise goes one step further: it combines wheel and rail acoustics roughness (its
generating mechanism) with vehicle and track transfer functions.
The derivation of these transfer functions is addressed in the standard. The complete process of an
environmental prediction noise scheme showing different types of inputs is illustrated in Figure 1. It
shows the different types of inputs: traffic data, acoustic parameters and geographical data.

Figure 1 — Elements of an environmental noise prediction scheme
1 Scope
This document addresses the measurement of source terms for environmental noise calculation for rail
traffic (including light rail, such as trams, metros, etc.). It is applicable to the measurement of in-service
trains on operational tracks.
It is not applicable to type acceptance testing of rolling-stock or tracks, or to derive source terms for time
domain models.
The following rail traffic noise source types are in the scope:
— rolling noise;
— traction and equipment noise;
— aerodynamic noise;
— impact noise (e.g. rail joints, switch and crossings, wheel flats);
— braking noise;
— bridge noise;
— squeal noise.
Noise from rail vehicles at standstill, such as stationary engine idling and auxiliary equipment at yards
and stations, is covered by EN ISO 3095:2013 for measurement procedures and operating conditions, and
by EN ISO 3740:2019 and EN ISO 3744 for the determination of sound power. It is therefore not in the
scope of this document.
The calculation of the propagation of sound is part of generally standardized propagation models which
are not addressed in this document.
Noise from fixed installations (e.g. stations, depots, electricity sub-stations) is not in the scope of this
document.
Source terms are specific to a vehicle and track type. The scope includes measurement procedures and
conditions and sampling requirements.
2 Normative references
The following documents are referred to in the text in such a way that some or all of their content
constitutes requirements of this document. For dated references, only the edition cited applies. For
undated references, the latest edition of the referenced document (including any amendments) applies.
EN 15610:2019, Railway applications — Acoustics — Rail and wheel roughness measurement related to
noise generation
EN 17343, Railway applications — General terms and definitions
CEN/TR 16891:2016, Railway applications — Acoustics — Measurement method for combined roughness,
track decay rates and transfer functions
EN IEC 60942:2018, Electroacoustics — Sound calibrators (IEC 60942:2017)
EN 61094-4:1995, Measurement microphones — Part 4: Specifications for working standard microphones
(IEC 61094-4:1995)
EN 61260-1:2014, Electroacoustics — Octave-band and fractional-octave-band filters — Part 1:
Specifications (IEC 61260-1:2014)
EN 61672-1:2013, Electroacoustics — Sound level meters — Part 1: Specifications (IEC 61672-1:2013)
EN 61672 (all parts), Electroacoustics — Sound level meters
EN ISO 3095:2013, Acoustics — Railway applications — Measurement of noise emitted by railbound
vehicles (ISO 3095:2013)
EN ISO 3744, Acoustics — Determination of sound power levels and sound energy levels of noise sources
using sound pressure — Engineering methods for an essentially free field over a reflecting plane (ISO 3744)
ISO 5725-2:2019, Accuracy (trueness and precision) of measurement methods and results — Part 2: Basic
method for the determination of repeatability and reproducibility of a standard measurement method
3 Terms and definitions
For the purposes of this document, the terms and definitions given in EN ISO 3095:2013, EN 17343, and
the following apply.
ISO and IEC maintain terminology databases for use in standardization at the following addresses:
— ISO Online browsing platform: available at https://www.iso.org/obp
— IEC Electropedia: available at https://www.electropedia.org/
3.1
number of axles
N
ax
number of axles in the selected train or part of train
3.2
third octave band frequency
f
centre frequency of a third octave frequency band in Hz
3.3
third octave wavelength
λ
centre wavelength of a third octave wavelength band in m
3.4
acceleration signal
a(t)
time signal of the rail acceleration in m/s
3.5
equivalent vertical rail vibration level spectrum
L (f)
aeq,Tp
third octave spectrum of the rail head acceleration energy averaged over pass-by time T , in
p
dB re 1 µm/s
3.6
rolling noise transfer function
L (f)
HpR,tot,nl
transfer function in third octave bands between the sound pressure at a fixed point, 7,5 m, and the
combined effective roughness frequency spectrum, normalised to the axle density
N /ℓ, in dB re 20 Pa/√m
ax
[SOURCE: CEN/TR 16891:2016 definition 3.14]
3.7
combined effective roughness
roughness function that excites rolling noise
[SOURCE: EN 15610:2019, definition 3.4]
Note 1 to entry: The combined roughness is the RMS of the rail and wheel roughness spectra. It becomes the
combined effective roughness when the effect of the contact patch filter is included.
3.8
combined effective roughness wavelength spectrum
L (λ)
Rtot
wavelength spectrum in third octave bands of the combined effective wheel-rail roughness including the
contact filter, in dB re 1 µm
[SOURCE: CEN/TR 16891:2016, definition 3.12]
3.9
combined effective roughness frequency spectrum at speed v
L (f,v)
Rtot
frequency spectrum in third octave bands of the combined effective wheel-rail roughness at a given speed
v, including the contact filter, in dB re 1 µm
[SOURCE: CEN/TR 16891:2016, definition 3.13]
3.10
direct roughness measurement method
refers to an acoustic roughness measurement method for which the sensor measures the running surface
roughness so that either the rail or the wheel roughness is measured independently of any effect of wheel-
rail interaction
[SOURCE: EN 15610:2019, definition 3.5]
3.11
indirect roughness measurement method
refers to an acoustic roughness measurement method that measures a quantity that is the result of wheel-
rail interaction, such as noise, rail or axle box vibration, whereby the original excitation by the combined
effective wheel and rail roughness is inferred
[SOURCE: EN 15610:2019, definition 3.6]
3.12
sound power level
L
W
ten times the logarithm to the base 10 of the ratio of the sound power, P, of a source to a reference value,
P , expressed in decibels, where the reference value, P , is 1 pW
0 0
[SOURCE: EN ISO 3740:2019, definitions 3.2 and 3.3]
Note 1 to entry: In this context, this is the sound power level as derived from a specific field measurement point. It
is the sound power level required to produce the sound pressure level in the measurement point at the trackside in
third octave or octave bands.
3.13
integration length
ℓ length of vehicle (buffer-to-buffer) or track over which the sound pressure is integrated, in m
3.14
sound power level per unit length
L
W'
sound power level L normalized to the length ℓ
W
L ’ = L – 10 lg (ℓ / ℓ ) in dB re 1 pW/m (1)
W W 0
where ℓ is 1 m
3.15
sound power transfer function
L
HWR,n
third octave transfer function of sound power per unit (combined) effective roughness, per axle, in one
third octave bands, in dB re 1 W/m , in analogy with [1], formulas 2.3.8 - 2.3.10
L (f) = L (f)– L (f,v) – 10 lg N in dB re 1 W/m (2)
HWR,n W R ax
Note 1 to entry: This depends also on vehicle length.
3.16
sound pressure transfer function
L (f)
HpR,nl
third octave transfer function of sound pressure at the trackside, per unit (combined) effective roughness,
normalised to the axle density N /ℓ
ax
0,5
L ,nl (f)= L (f) - L (f,v)– 10 lg (N /ℓ) in dB re 20 Pa/m (3)
HpR peq,Tp R ax
[SOURCE: CEN/TR 16891:2016, definition 3.14]
Note 1 to entry: In principle, this is independent of the train or vehicle length and speed if rolling noise is the main
source.
3.17
acoustic transfer function from sound pressure to sound power
L (f)
HpW
transfer function between the measured pass-by sound pressure level L and the sound power level
peq,Tp
per meter L , in one third octave bands or octave bands
W’
Note 1 to entry This only includes geometric divergence and ground attenuation and results from integration over
the pass-by of single or multiple sound sources for a vehicle or group of vehicles.
3.18
distribution function
D(f)
set of two or more frequency spectra in one-third octave bands, used to quantify the contributions of
individual sources, such as vehicle and track, or rolling noise and other sources
Note 1 to entry: The energy sum of these spectra is always 0 dB in each one-third octave band. The contribution of
each source is obtained by subtraction of the distribution function from the total sound spectrum.
3.19
unweighted pass-by sound pressure level
L (f)
peq,Tp
unweighted sound pressure level integrated over pass-by time T , in one third octave bands,
p
in dB re 20 µPa
3.20
source term
sound power level per unit length either for a vehicle and track or for a traffic flow, as used in
environmental noise prediction models
Note 1 to entry: For each vehicle and track, source terms are defined at specific heights and represent different
physical sources (e.g. rolling noise, traction noise and others).
Note 2 to entry: Source terms for a whole traffic flow are based on the mix of trains and their speeds at a given site.
4 Instrumentation and calibration
4.1 General
Measurements can be taken with temporary, supervised setups as well as with stationary mounted,
automated, non-supervised setups. While most instrumentation and calibration requirements are
necessary for both types of setups, some extra requirements are necessary for stationary ones.
Furthermore, some tasks conducted by the supervisor for temporary setups also have to be executed by
instrumentation for stationary mounted setups.
Depending on the type of measurement, only a selection of the following instrumentation might be
needed.
4.2 Acoustic instrumentation
The microphones, signal acquisition units and processing algorithms used shall each comply with the
requirements of EN 61672-1:2013 specifications for class 1 measuring equipment.
Where measurement equipment other than type-approved sound level meter is used, microphones shall
comply with the requirements of EN 61094-4:1995 specifications for class 1 measuring equipment.
NOTE Multichannel acquisition systems are generally used to record data.
In the case of measurements of survey grade, this requirement is relaxed to class 2 instruments.
The sound calibrator shall meet the requirements of class 1 according to EN IEC 60942:2018.
Microphones with free-field or diffuse-field characteristics shall be used. A suitable microphone
windscreen should be used. Stationary mounted setups shall be suitable for permanent outdoor usage
and shall include bird spikes.
Where one-third octave frequency band analysis is required, the filters shall meet the requirements of
class 1, according to EN 61260-1:2014.
The conformity of the calibrator with the requirements of EN IEC 60942:2018 shall have been verified
within one year of the test date. The conformity of the instrumentation system with the requirements of
the EN 61672 series shall have been within two years of the test date. The date of the last verification of
the conformity with the relevant standards shall be recorded.
4.3 Calibration of the acoustic instrumentation
4.3.1 Temporary setup
Before and after each series of measurements taken with a temporary setup, a sound calibrator shall be
applied to the microphone(s) to verify the calibration of the entire measuring system at one or more
frequencies over the frequency range of interest. If the difference between two consecutive calibrations
is more than 0,5 dB, or the difference between a measured value and the nominal value is more than
1,1 dB, all of the measurement results in between shall be rejected.
The sensitivity of the measurement chain actually applied in the field shall be documented.
4.3.2 Automated calibration
For stationary-mounted setups, the manual test with a calibrator, as described above, shall be conducted
at least once a year. Additionally, automated daily tests shall be conducted by one of the two following
procedures.
Testing the instrumentation by supplying a known electric signal. The difference between the measured
value and the nominal value shall be not more than 1,1 dB. This can be done by built-in acoustic sources
in the microphone (signal at the microphone output has to be equal to a signal of at least 80 dB), charge-
injection-calibration, electrostatic calibration or any other equivalent procedure.
Testing the instrumentation by additional measurements (other microphones or sources). For this, a
second microphone is installed. The difference between the measured values of both microphones shall
not exceed 1,5 dB.
If the daily test of the instrumentation fails, all of the measurements taken since the last successful daily
test shall be rejected.
4.4 Non-acoustic instrumentation
4.4.1 Time and duration
The date and local time of each pass-by shall be recorded. The duration of the pass-by of the considered
train/unit (over buffers) shall be measured.
The pass-by time T can be determined from the front and rear axle peaks in the time signal of rail or
p
sleeper vibration or sound pressure, plus a time increment proportional to the distance of the first and
last axle to the front or rear of the train respectively. Alternatively, an optical presence detector can be
used to directly determine T .
p
4.4.2 Speed, track and direction
The track number used by the train together with the direction of travel shall be recorded. The speed of
the train shall be measured during the pass-by with a precision of better than ±3 km/h for velocities up
to 100 km/h and better than 3 % of the measured value for velocities above 100 km/h.
4.4.3 Meteorological parameters
Weather conditions influencing the measurements shall be recorded, including temperature and wind
speed. Where possible, rail temperature can also be registered.
NOTE Temperature can affect the rolling noise level.
4.4.4 Combined roughness
The combined roughness shall be measured according to the requirements in CEN/TR 16891:2016.
Figure 2 shows under which circumstances this measurement is necessary.
In this case track decay rates and a threshold value for rail roughness can be obtained from rail vibration
of multiple pass-bys.
4.4.5 Wheel and track parameters
Where the track decay rate is required, it should be measured within 1 year before or after the
measurements according to the requirements in EN 15461. If direct rail roughness is to be measured, this
should be in accordance with EN 15610. Figure 2 shows under which circumstances this measurement is
necessary. Rail roughness measurements should be made, where possible, within 3 months before or
after the pass-by measurements. Sound and rail roughness measurements should be made a sufficient
period after grinding, typically around three months, to ensure that rail roughness has stabilized.
If direct wheel roughness is to be assessed, it shall be measured in accordance with EN 15610.
NOTE Wheel roughness immediately following reprofiling might not be representative of normal operation.
5 Approach to derive source terms
5.1 General
The sound power, used for source terms in prediction models, is a measure of the strength of a source
suitable for estimations of environmental noise in terms of long term equivalent sound pressure level
arising from a flow of traffic along a route. A sound power is an estimate calculated from sound pressure
level measurements made at a limited range of angle from the source and makes use of a directivity index
(DI) assumed for all sources of similar type by the convention of a particular environmental noise
calculation scheme. The definition of a sound power therefore includes a statement of the DIs assumed
in its estimation from measurements. They can therefore differ depending on the prediction model.
The symbol L is used here for the sound power, taking into account that it has a different definition for
W
specific prediction models.
NOTE There is no attempt to integrate the actual sound power over an enclosing surface and no normalization
of the assumed directivity to account for the complete sound power of the source. The assumed sound power is
therefore not directly related to the true sound power of the source.
As source terms are derived from one or more microphone positions, propagation back to the required
source height is done by means of a transfer function L , which is either tabulated or calculated.
HpW
This clause therefore sets out the requirement for estimating a sound power source spectrum from the
sound pressure measurements.
5.2 Procedure
The procedure to derive the sound power level from sound pressure is as follows.
Measure L for each vehicle (i.e. each separate coach or wagon) at either the 3,5 m or 1,2 m
peq,Tp
microphone height above the top of rail and at 7,5 m from the centreline of the track. If the sound sources
on one side of the train are stronger than the other, the measurement from that side should preferably
be used. In order to separately measure unique sources such as cooling fans or engines, conditions are
preferred at which these are predominant, for example at lower speeds or on tracks with low rail
roughness.
Measurement shall preferably be over a flat ballast layer as described in 7.2.
NOTE The ground attenuation effect can be determined with greater certainty if there is a ballast shoulder and
known ground geometry beyond the ballast shoulder.
Measurements shall be taken by any of the following options, depending on uncertainty requirements
and practical considerations:
For the least uncertainty, quality A, the microphone position at 3,5 m height is preferred. The uncertainty
in the step of estimating L from L is evaluated as σ = ±3 dB for 0,5 m source height, ±2 dB for 1 m, 2 m,
W p H
or 4 m source heights. For trams with shielded wheels and undercarriage, the lower microphone position
of 1,2 m height is acceptable and the same uncertainty can be assumed.
For the 1,2 m microphone position, which results in greater uncertainty, results shall be reported as of
quality B, with uncertainty of σ = ±4 dB for all source heights in any one-third octave band.
H
If measurements are taken at sites of other geometry than over flat ballast (see 7.2), calculation of L
HpW
is required. The uncertainty is then undefined by the standard, and shall be reported as of quality C, along
with the method applied. In that case, the uncertainty is expected to be smaller than in cases A and B if
L is calculated for the exact geometry of the test site.
HpW
The measurement should be repeated for at least 3 trains or rail vehicles of the same type travelling at
speeds within ±8 %. [8 % is 1 dB following 30 log10(v)]. The standard deviation, σ of the measured
M
levels of L shall be calculated for the total A-weighted level.
peq,Tp
The sound power L (f) shall be estimated as a one-third octave spectrum as
W’
L ’(f) = L (f) – L (f) (4)
W p HpW
The transfer function L is obtained by any of the following:
HpW
— From tables provided in Annex B, depending on site geometry and source/receiver positions,
(only for receivers at 7,5 m from the track centreline) with uncertainty of quality A or quality B.
— Calculated with an appropriate model, taking direct, indirect and diffraction paths into account (see
Annex B), with uncertainty of quality C
— Using transfer functions derived from the prediction model concerned. In some cases however, site
effects will not be properly taken into account, resulting in an uncertainty of quality B or larger.
The uncertainty in the measurement of the sound power spectrum shall then be calculated as
2 2
σ = √(σ + σ ). (5)
W H M
The method applied should be reported including uncertainty.
5.3 Directivity
Directivity of sources is included in prediction models (such as CNOSSOS) and can differ between these
models. The transfer function from sound pressure to sound power L includes the directivity in the
HpW
integration over the source pass-by. The tables given in Annex B specify the directivity for different cases.
5.4 Distribution over source heights
The following principles are applicable to the distribution of sound power over two or more source
heights.
The energy sum of contributions from all sources at all heights should be the same as the total noise at
the measurement position(s).
The allocation to each source height should depend on the knowledge of the position of the sources. The
spectral contribution of each source can be based on known characteristics of the sources.
If required, a microphone array can be used specifically to determine the contributions at the defined
source heights, for example by mapping the sound power over each source area associated with each
source height. The procedure used should be referenced or described in the report. Examples are shown
in [8].
6 Measurement procedures
6.1 General
The choice of measurement sites will depend on the type of source, vehicle operating conditions, and the
available speed range at each site. But also, practical considerations shall be taken into account, such as
accessibility and representativeness.
Several measurement sites can be required to allow collection of sufficient data from in-service trains at
different speeds. Site requirements are set out in 7.2.
As the measurement effort for source strength can easily become too large, it is important to allow
elimination of sources if it can be demonstrated that they are not relevant or typical. Before
characterizing each individual source, an evaluation should be made of which noise sources and
operating conditions are relevant for the vehicle in question based on knowledge of the vehicle and its
design, and/or available measurements. Sources that do not exceed the others by more than 3 dB in one
or more third octave frequency bands can be omitted. This can for example occur for tonal traction noise,
braking or curve squeal noise.
An overview of each source type, typical speed range and other characteristics is given in Table 1,
together with reference to the relevant clause. This table can be used to prioritize which sources to
include.
Table 1 — Overview of railway noise sources and their typical speed range and characteristics
Source Indicative speed Notes
range
Rolling noise (6.2) > 10 km/h Generally dominant source on plain line track
Traction and equipment 0 - 80 km/h Particularly for acceleration, low speeds and for
noise (6.3) slopes
Impact noise (6.4) > 10 km/h Typical for jointed track, switches, track defects
and wheel flats.
Curve squeal (6.5) > 10 km/h Typically when curve radius is below 150 m.
Rolling stock dependent.
Bridge noise (6.6) > 10 km/h Characterizes additional noise radiated by
bridge or viaduct structure. Typically significant
for steel structures and close to or underneath
bridges.
Braking noise (6.7) All speeds Can be significant for approaches to stations.
Aerodynamic noise (6.8) > 250 km/h Typically significant source at high speeds.
Can also be significant below 500 Hz at lower
speeds. Increases at higher rate than rolling
noise with speed.
Due to the many potential uncertainties in source data collection, it is recommended to validate the
resulting source terms by feeding them into the relevant prediction model and comparing the results
with measurements at other sites.
In Figure 2 a flow diagram of the various steps in the process to determine source terms is shown, also
indicating the relevant section in this document for each. Different paths are followed for prediction
models that use transfer functions, than those that do not.
The required effort depends on whether new source data are determined, or only a comparison is
performed to assess whether a new train type fits into an existing category.
Key
Planning requirements
Measurements
Measurement and processing for transfer functions
Processing and analysis
Determining source terms
Figure 2 — Flow diagram of process to determine railway source terms, with relevant section
numbers indicated
6.2 Rolling noise
6.2.1 General
Rolling noise is radiated by structural vibration in rails, sleepers, wheels and in some cases (mainly some
types of freight wagon) also by the vehicle superstructure. It is generated by the combined wheel and rail
roughness.
For the CNOSSOS prediction model as described in Directive (EU) 2015/996 [1], the source term
quantities are listed in 2.3 of the Directive and some default values are given in Appendix G, tables G1, G2
and G3. Default values for Harmonoise/Imagine can be found in [6].
6.2.2 Measurement steps
The rolling noise sound power spectra from vehicle and track can be most generally calculated using the
combined effective roughness and vehicle and track transfer functions. These are obtained from
measurements carried out for each specific train type and track type in the following steps:
a) Select train speeds and operating conditions such that rolling noise is ensured to be the dominant
source (see 7.3). Sites with smooth rail roughness are preferable, unless a high rail roughness is
characteristic for the network.
b) Determine the combined effective roughness L (f), applicable to the measurement site and vehicle
Rtot
type, as a function of frequency, in one-third octaves for each pass-by at speed v as set out in Annex A.
Further details are given in CEN technical report CEN/TR 16891:2016, Clause 13. Alternatively,
combined roughness can be determined from the directly measured wheel and rail roughness.
c) Measure the sound pressure L (f) in one-third octaves for each pass-by at speed v;
peq,Tp
Determine the total transfer function L (f) from combined roughness and sound pressure as:
HpR,nl,tot
L (f) = L (f) – L (f,v)- 10 lg(N /ℓ) (6)
HpR,tot,nl peq,Tp Rtot ax
NOTE 1 This total transfer function is independent from the roughness, speed, train length and number of
axles, if only rolling noise is present. It characterizes the vibroacoustic properties of the vehicle, track and the
propagation area.
d) Average the total transfer function arithmetically for the speeds measured;
If individual results for the transfer function differ from others by more than 5 dB in single third
octave bands, then a larger number of measurements should be taken, or individual measurements
rejected.
e) Calculate the total sound power transfer function L (f) from L (f) according to Annex B.
HWR,tot,nl HpR,tot,nl
f) Determine the vehicle transfer function LHWR,VEH,nl(f) and track transfer function LHWR,TR,nl(f) by
applying separation to the total transfer function L (f) as described in Annex C.
HWR,tot,nl
Steps c) and d) to determine the transfer function L of combined roughness to total sound pressure
HpR,nl
normalized to the axle density, N ℓ are in accordance with CEN/TR 16891:2016, Clause 14. This can be
ax/
converted to the transfer function of combined roughness to sound power per axle L (f) for given
HWR,n
vehicle length ℓ or normalized to axle density N/ ℓ, L . For the simple case of a cylindrical source
HWR,nl
along the track centreline with measurement distance r:
L (f) = L (f)+ 10 lg (2πr) (7)
HWR,n HpR,nl,
for the free field term (geometric divergence),
and
L (f) = L (f) - L (f) (8)
...