Headed for the Low-Noise Railway:

The DB Noise Reduction Research Programme

B. Schulte-Werning, M. Beier, H.-P. Grütz, K. Jäger, G. Kock, J. Onnich, R. Strube

Deutsche Bahn AG,
Forschungs- und Technologiezentrum
Völkerstr. 5, D-80939 München
Tel: +49 89 1308 2581, Fax: +49 89 1308 2491
Email: [email protected]

ABSTRACT
The research programme of the Deutsche Bahn AG (DB) is presented which aims for a major
leap within the next years towards less railway noise production.
In five projects primary noise experience via the ambient air as well as the secondary noise
experience via the ground vibrations are tackled from the source to the residents. Beside these
investigations, the sound perception and noise assessment is paid attention, too.

1. - INTRODUCTION
In the last few decades, mobility in Europe, based on automobiles for individual traffic and
lorries for freight traffic, has grown more and more; yearly growth rates of more than 10%
have often been reported by statistics. These growth rates cumulated over the years cannot
easily be handled in an environment-friendly way. The European Commission has therefore
given distinct political signals to get more passengers and more freight onto the railways. An
efficient trans-European transportation network is indispensable to handle this mobility in-
crease.
To empower the environmental friendliness of the railway traffic in Germany [13], the
Deutsche Bahn (DB) is in between of absolving a major research programme concerning
noise reduction. The final goal of this ambitious research agenda is to achieve a decline in
overall noise production by up to 10 dB(A) compared to 1990 by the year 2005 for DB's busi-
ness units for freight and passenger traffic as well as the infrastructure division.

2. - NOISE REDUCTION RESEARCH AT DB

The representative "acoustical" situation around a train in motion on a railway line is shown
in Figure 1. Train and track both act as the sound sources and the two ways of noise transmis-
sion make two lines of research indispensable.

Figure1: Acoustical situation around a train in motion on a track section

Since in the air only longitudinal waves are possible, the classical rules of acoustics can be
used to trace the sound waves from the source to the railway line residents whom they may
disturb (see e.g. [10]). The direct noise experienced by these residents can be reduced both by
limiting the strength of the sound source and by extending the transmission path of the sound
wave by making use of the screening effect of sound barriers. Also the installation of noise
protection windows represents a counter-measure against too high noise exposure.

In the ground, the transmission of energy is possible by longitudinal as well as by transversal

waves and different mixed wave structures can be identified apart from these pure wave types.
This makes the analysis of the acoustical problem from the source to the residents' ears more
complex since transmission of the wave energy through the ground and formation of the sec-
ondary sound in the houses involve more parameters and influencing quantities.

To realise this goal, seven projects are running within the DB programme "Low Noise Rail-
way". These deal simultaneously with the noise treatment of the trains and the rail/wheel sys-
tem and include vibration control, the reduction of noise transmission, the assessment of noise
perception as well as the upgrading of the specially monitored track technique (Figure 2). The
support of the revision of the two German legislative acoustic-guidelines SCHALL03 for
regular passenger and freight traffic along with AKUSTIK04 for shunting yards and tran-
shipment terminals complete the efforts. External scientific institutions, engineering experts
and manufacturers are involved in all of the projects in order to obtain a broad view of the
problems stated.
All these projects are supervised on a regular basis by associated steering committees with
members of the DB’s business areas. They support both the project leader during the work
progress and the integration of the results into the daily business as early as possible.

DB Research Programme "Low Noise Railway"

Noise Emission of Trains Vibration and

and Wheel/Track System Secondary Sound

Noise Emission Sound Perception and

of Railway Line Noise Assessement

Noise Cutback Revision of Guidelines

by Sound Barriers SCHALL 03 and AKUSTIK 04

Upgrade of the
Specially Monitored Track

Figure 2: Noise Research Programme of Deutsche Bahn AG (DB

Thus all impacts of rail-bound vehicle noise, from the sources of noise to the residents near
the railway line, are being tackled. Overall expenditure per year is approximately 2.5 million
Euro which shows the high priority of these noise reduction activities.
 2.1 - NOISE EMISSION OF TRAINS AND WHEEL/TRACK SYSTEM
Since the minimisation of sound radiation directly at the source is the first-choice solution,
several work packages deal with the noise emission of the trains and the wheel/track system.

The necessary understanding of the sound sources of high-speed trains was gained within the
French-German DEUFRAKO framework in the project K2 "Noise sources from High-Speed
guided Transport" [8]. Combined with the DB work in the acoustical analysis of freight traffic
in the EU-funded projects SILENT FREIGHT / SILENT TRACK, see e.g. [11], the basis was
found for the progressive achievement of low-noise traffic in Europe. Currently the work for
the EU-funded project HARMONOISE and the “Leiser Verkehr”, funded by the German
government, is performed.

Railway coaches and wagons are equipped with up to 8 wheels and the discs and rims of these
wheels are excited by wheel/rail interaction and dominate overall sound radiation. Since this
high "loudspeaker" area per unit train length cannot be reduced e.g. because of axle load
limitations in freight traffic, silent railway wheels - both for block and disk brake operation -
will play an important role in the future to meet the goal of a low-noise railway system.
The basic studies concerning the optimum shape with respect to the acoustic requirements
have already been performed (Figure 3), see e.g. [4], [14]. A wheel manufacturer as well as
wheel and brake engineers are involved in the workgroup to clarify the questions of the cast-
ing process and of the brake and running safety. The full scale investigations will start in
brake test bench in autumn 2001 and will be finalised in 2002 with an in-situ test under real
operational conditions.

Figure 3: Numerical simulation of the sound radiation of railway wheels with different shapes

Within this project, the development of DB’s microphone array to be equipped with up to 96
microphones was done and first measurements were performed. The array characteristics,
resolution and signal-to-noise ratio were evaluated by numerical simulations [16], [17]. Then,
array measurements were used to separate wheel and rail noise from a high-speed train. With
this array, DB is now able to localise the dominant noise sources in the relevant frequency
bands [6], [20].

Figure 4: Noise source localisation of microphone-array measurements

Special attention will be paid in the future to the noise radiation from the pantographs on the
engine roofs, as well as to other noise sources which may be located high above the ground,
thus being difficult to screen by barriers [1], [2], [3].

2.2 - NOISE PRODUCTIO N FROM RAILWAY LINE

In the last years, DB has implemented the acoustical concept of the "Specially Monitored
Track (SMT)" (in German: "Besonders überwachtes Gleis (BÜG)"). The SMT concept is
based on the periodic acoustic monitoring of the relevant track section by means of a test
coach specialised for sound measuring. Should noise production be beyond a certain limit,
grinding of the track section will remove the tiny periodic rail corrugations which excite the
train wheels and lead to sound radiation from the wheels and the rail itself. Figure 5 shows a
representative example of the sound pressure level before and after grinding the rails.

Figure 5: Noise emission from railway line before and after local grinding

The Federal Railway Office (EBA) officially confirmed SMT as a noise reduction system
with effectively -3 dB(A) for ballasted tracks and slab tracks.
Further development of the standard SMT process was done towards a mobile measuring de-
vice for rail surface roughness to perform in-situ quality control directly after grinding and
towards an application of SMT to slab tracks equipped with a sound absorbing layer.

To complete these activities, an automatic system for sound diagnosis was developed to
measure the sound production of the different train fleets at a given track location over a
longer time period to gain a better insight into the scatter in the noise emission within a this
fleet and its impact on the line emission.

2.3 – UPGRADE OF THE SPECIALLY MONITORED TRACK

The experience of the last years now showed, that the authoritatively stated compatibility
conditions for the use of the SMT, i.e.
periodic acoustic monitoring of the relevant track section by means of a test coach spe-
cialised for sound measuring
monitoring of the grinding quality of the ground track section
application of the different grinding procedures
need further development and have to be adjusted to the technological progress.
DB’s research is therefore marching in two directions simultaneously: to reduce the expenses
for this particular reduction rate and to increase for the current cost structure the noise reduc-
tion rate.

The typical noise reduction curve over time of the current SMT in Germany is shown in Fig.
6. If a certain track section is subject to the SMT, grinding will immediately decrease the rail
noise clearly below 48 dB(A), this value is identical with the so called “base value” in the
German legislative framework SCHALL 03. Approximately half a year later the track section
emits an even 1 to 2 dB(A) lower sound pressure level. Afterwards the noise emission in-
creases with time and after around 2 years the so called threshold value of 51 dB(A) is
reached, this initiates the next rail grinding. In the current practice the noise emission oscil-
lates over time around the mean value of 48 dB(A) with an amplitude of 3 dB(A).

Figure 6: Noise reduction grinding as a function of time of the SMT technique using track
grinding

The idea behind the “low-cost” SMT option is now to follow the base value by increasing the
frequency of the grinding with a decreased grind-amplitude. The current research is now
looking for the cost optimised frequency-amplitude pairing.
The “high-performance” SMT option is working with an even higher frequency and smaller
amplitude of the grinding in comparison to the low-cost option, but the striking feature is the
lower base value. This option aims for a noise reduction rate of effectively –5 dB(A).

Work on rail corrugations which cause rail noise and also excite the wheels is currently fo-
cussed on the analysis of the measured data concerning these corrugations and on tracing the
growth rate of the corrugations on track sections with different grinding procedures [5].

2.4 - NOISE REDUCTION BY MEANS OF SOUND BARRIERS

In relation to rolling noise, the typical screening effect of a 2 m high sound barrier is average
10 dB(A). At up to 1 million EURO/km, the installation of sound barriers is costly and to jus-
tify these substantial investment costs the effectiveness of the noise protection installations
must be as high as possible. Barriers with 3 or 4 m height will be an exception since they im-
pair the residents. Furthermore, train passengers can no longer enjoy the outside landscape
through the coach windows and travelling by train becomes less of a pleasure.

As sound inflection reduces the protection effect of the sound barriers, another means to in-
creasing the effectiveness of the barriers is a better understanding of the inflection properties
followed by systematic optimisation of the construction of the wall. Since the inflection effect
is dominated by the structure of the wall edge, the numerical simulation (Figure 7, see [14]) of
several detailed configurations will provide an initial selection of the configurations for later
full-scale testing.
The in-situ tests of a “hand-made” prototype showed that the principle of the acoustically soft
edges works very well and that the barrier effect can by improved by several dB(A) depend-
ing on the distance to the track and the height. Advancement is on going to a prototype ready
for an industrial manufacturing process while preserving the acoustic quality.

Figure 7: Simulation of sound cutback effect by barriers with varying inflection properties

These investigations are being completed by the quantification of the damping effect of the
ground and meteorology in between the sound source and the residents to optimise the overall
noise protection. Furthermore the noise damping degree of barriers made of glass are to be
investigated, this kind of barriers come into fashion as they are no optical impairment to the
residents. These results will also enter the sections related to noise protection in the environ-
mental protection laws in Germany.

2.5 - SOUND PERCEPTION AND NOISE ASSESSMENT

The description of a certain sound experience as annoying "noise" is a very individual judge-
ment and can by no means be considered in terms of physical quantities alone. The question
of when sound is qualified as noise is therefore a very basic one and it is investigated parallel
to the above-mentioned activities taking physiological and psychological aspects into account.

In several European countries, there is a "bonus" approach towards railway-created noise

(Corrective factor rail/road traffic noise). This is based on consideration of the fact that, for a
given hourly passing according to sound pressure level, railway noise is to some extent less
annoying than noise created by private car and lorry traffic because of the dissimilar sound-
production characteristics. The basic investigations were done in the 80s, leading in Germany
since 1990 to the permitted “rail bonus” of -5 dB(A) [19]. The same holds for Austria, while,
for example, in France it is -3 dB(A) and in Switzerland it is -5 dB(A) to -15 dB(A) depend-
ing on the intensity of the rail traffic.

In the meantime the railway traffic situation in Germany changed significantly as e.g. high-
speed trains came into operation, existing lines were upgraded for operation up to 200 km/h
and the mix of passenger and freight traffic shifted. To reflect these changes and to support
the legislative authorities, the assessment of noise quality is re-investigated in detail. The fo-
cus is on the comparison of rail and street noise, the perception of high-speed trains noise and
the annoyance of conventional passenger and freight train noise. Also the alteration in the
degree of annoyance near new built and upgraded railway lines were tackled.
The investigations already finalised show clearly, that the rail bonus is still a valid approach
for the annoyance disparity between railway and street traffic and that –5dB(A) are a least
gratuity all over the day (Figure 8). A novel outcome was the variation in the annoyance dis-
parity during the day, in the night-time the bonus could be even much higher. Furthermore the
high-speed traffic is not more bothering to the residents than standard passenger trains [9],
[12], [15], [18].

Figure 8: Compacted result of the noise assessment studies illustrating less annoyance to
residents by the rail traffic compared to street traffic)

The avoidance of the wheel squeal to be heard from freight wagons on very small-radius track
curves is a distinct goal. Up to now, the occurrence of this noise component in freight-train
shunting yards could only be reduced by costly sound barriers.
To improve this situation, research was performed into avoiding the noise at the source. Sev-
eral techniques were tested among which the suitable artificial lubrication of the wheel-rail
surface was successfully leading to a noise reduction. The safety of the train operation is not
affected by this method, as the braking system is in essence designed for the “natural” lubri-
cation by the autumn leafs. Currently the approval of the federal authorities is aimed for.

2.6 - VIBRATIONS AND SECONDARY SOUND

Train movement on the track not only causes direct audible sound via transmission through
the air but also sets off vibrations which are transmitted through the soil. Away from the rail-
way line, these vibrations may reach a building thus creating both vibrations in the structure
itself and also so-called secondary sound in the low frequency range due to the excitation of
the walls and floors in the building.
At present legal specifications for tolerable vibration limits do not exist in Germany. Even so,
some industrial standards and norms exist. In combination with the difficulty of forecasting,
this results in some uncertainty concerning the vibration values to be aimed for during the
planning phase of construction projects for various new railway lines or for upgrading exist-
ing lines. More or less reasonable error margins are taken into account to be "on the safe side"
concerning later complaints by residents.

To overcome this uncertainty, the validation and application of numerical simulation models
for vibrations is essential. The scope is wide-ranged in the sense that, from the train/track in-
teraction through the transmission to the formation of the secondary sound in the house near
the railway line, the simulation should give a reliable forecast. Also the effect of vibration
protection walls buried in the ground has to be calculated with precision. Beside the physical
modelling, the input information such as the material data of the ground is essential for the
success of the simulation.
As an example, Figure 9 shows such an approach based on a complex spring-damper model-
ling of a track section creating ground vibrations during train passage. This physical model is
used in the DB computer code RIM-Ground [7].

Figure 9: Physical model of a track section creating ground vibrations during train passage
in the RIM-Ground computer code

The definition of suitable vibration limit values for residents is, of course, also vital to allow
for more confidence in the planning procedures. Here, laboratory investigations with volun-
teers exposed to vibrations of the "railway-induced type" are underway and will form a data
basis concerning the perception of, for example, the average and peak values and their spec-
tral composition.

The results are used to set up a DB-internal guideline concerning the vibration protection e.g.
for new lines under construction. But also the situation at existing lines is tackled. This
guideline assists the civil engineering during the planning and construction phase of the infra-
structure projects.
 2.7 - REVISION OF GUIDELINES SCHALL03 AND AKUSTIK 04
The German acoustic-guidelines SCHALL03 (calculation rules for sound receiving near rail-
way lines) and AKUSTIK04 (calculation rules for sound receiving near freight-train shunting
yards and transhipment terminals) were released in 1990 by the legislative authorities [19].
Figure 10 demonstrates the basic physical model of the SCHALL03 for the calculation of the
noise received by the residents. Both guidelines reflect measurements and investigations per-
formed in the 80s.

Figure 10: Physical modelling of sound receiving near railway lines according to SCHALL03

In the meantime the railway traffic situation in Germany changed notably as among other
modifications the high-speed trains came into operation which can only be tackled insuffi-
ciently by the existing guidelines. On the other hand the progress in the field of railway-
acoustics research should enter the guidelines to keep them on a “state of the art” level both
for the professional users and the residents near the railway lines.

To reflect this reasoning and to revise both guidelines the German federal authorities, lead by
the ministry of transportation, are putting together a broad work-group consisting of environ-
mental institutions, governmental as well as non-governmental, acoustic-engineering offices
and (railway) transport undertakings. The acoustic specialists of DB were asked to enter this
group with their expertise, too.

3. - CONCLUSIONS
DB is undertaking a major noise reduction research programme to answer the public demand
for low-noise railway traffic in future. All aspects of the creation, the transmission and the
perception of noise are being tackled in several projects by various workgroups launched with
DB experts along with acoustic professionals from other European railway undertakings and
external institutions.

4. - REFERENCES
[1] K.-H. Althammer, W. Baldauf and T. Lölgen: Considerations for High-Performance
Pantographs, World Congress on Railway Research WCRR 1999, Tokyo, Japan, 1999
[2] W. Baldauf, R. Blaschko, W. Behr, C. Heine and M. Kolbe: Development of an
actively controlled, acoustically optimised single arm pantograph, World Congress of
Railway Research WCRR 2001, 25-29 November, Cologne, Germany, 2001

[3] W. Behr, T. Lölgen, W. Baldauf, L. Willenbrink, R. Blaschko, K. Jäger and

J. Kremlacek: Low Noise Pantograph ASP - Recent Developments; INTERNOISE
2000, 27-30 August, Nice, France, 2000

[4] W. Behr, L. Willenbrink, K. Degen, M. Beier, C. Katzenschwanz and D. Stüwing:

Modelling of wheel sound emission by computer-simulation, World Congress of
Railway Research WCRR 2001, 25-29 November, Cologne, Germany, 2001

[5] M. Beier: Short-pitch corrugation and noise emission, World Congress of Railway
Research WCRR 2001, 25-29 November, Cologne, Germany, 2001

[6] K. G. Degen, A. Nordborg, A. Martens, J. Wedemann, L. Willenbrink and M. Bianchi:

Spiral array measurements of high-speed train noise, INTERNOISE 2001, 27-30
August, The Hague, Netherlands, 2001

[7] R. Diehl, M. Beier, G. Hölzl and H. Waubke: Prediction of Railway induced Ground
Vibration, INTERNOISE 2000, 27-30 August, Nice, France, 2000

[8] DEUFRAKO K2: Noise Sources of High-Speed Trains, Joint French-German

Research Collaboration, 1996-1999

[9] B. Griefahn., U. Möhler and Schuemer, R. (Hrsg.): Vergleichende Untersuchung über

die Lärmwirkung bei Straßen- und Schienenverkehr (Aufweckstudie), München:SGS,
1999.

[10] M. Heckl and H.A. Müller: Taschenbuch der Technischen Akustik, 2. Auflage,
Springer-Verlag, Berlin, 1994

[11] B. Hemsworth, P.-E. Gautier and R. Jones: Silent Freight and Silent Track Projects,
INTERNOISE 2000, 27-30 August, Nice, France, 2000

[12] K. Jäger and J. Onnich: Fortschritte und Besonderheiten bei der Reduzierung des
Schienenverkehrslärms, Zeitschrift für Lärmbekämpfung 47, Nr. 6, November 2000

[13] M. Jaecker: Strategies of the German Government for the Reduction of Railway Noise
and their European Context, INTERNOISE 2000, 27-30 August, Nice, France, 2000

[14] T. Lölgen: Numerical Simulation of Railway Noise Generation and Propagation at the
Deutsche Bahn AG; World Congress of Railway Research WCRR ’97, Vol. E, pp.
315-322, Florence, Italy, Nov 1997

[15] U. Moehler and M. Liepert: Differences in the Annoyance between Rail and Road
Traffic Noise in Relation to the Acoustic Situation, INTERNOISE 2000, 27-30 August,
Nice, France, 2000

[16] A. Nordborg, A. Martens, J. Wedemann and L. Willenbrink: Wheel/rail noise separa-

tion with microphone array measurements, INTERNOISE 2001, 27-30 August,
 The Hague, Netherlands, 2001

[17] A. Nordborg, J. Wedemann and L. Willenbrink: Train noise source localisation with
microphone arrays, INTERNOISE 2000, 27-30 August, Nice, France, 2000

[18] R. Strube and M. Greven: Research on noise effects, World Congress of Railway
Research WCRR 2001, 25-29 November, Cologne, Germany, 2001

[19] 16. Verordnung zur Durchführung des Bundes-Immissionsschutzgesetzes

(Verkehrslärmschutzverordnung - 16. BImSchV), 1990

[20] L. Willenbrink, A. Nordborg, J. Wedemann and K. Degen: Localisation of train noise

sources by spiral-array measurements, World Congress of Railway Research WCRR
2001, 25-29 November, Cologne, Germany, 2001