1

ECS

www.esveld.com

Innorail2013

NOISE & VIBRATION MITIGATION IN RAILWAY TRACK

Coenraad EsveldEsveld Consulting Services

Emeritus Professor of Railway Engineering TU Delft

2

ECS

www.esveld.com

Innorail2013

Noise: Rolling; Engines; Curves; Braking; Aerodynamics.

Vibration: Rayleigh waves (at surface); Compression waves (tunnels); Shear waves (tunnels).

3

ECS

www.esveld.com

Innorail2013

CLASSIFICATION NOISE ANNOYANCE

4

ECS

www.esveld.com

Innorail2013

SQUEALING NOISE

Mitigating measures:• Lubrication;• Asymmetric rail grinding, shift of contact point wheel rail.

5

ECS

www.esveld.com

Innorail2013

CONTRIBUTION OF NOISE SOURCESTotal

Sleeper

Wagon

Rail Wheel

Sleeper

6

ECS

www.esveld.com

Innorail2013

METHODS OF NOISE REDUCTION Track Design

– Component selection (pads/sleepers)

– Embedded rail/special slabtracks– Damping

Barriers– Up to 10dB but affected by layout

of tracks/buildings– Expensive & visual impact– Low barriers & shrouds: not

interoperable

Acoustic grinding– Effective for corrugation– Not so effective for tracks in good

condition Absorptive Layers

– Low results on the rail– Higher results on slab track -

absorbs wheel noise Vehicle Design

– Wheel diameter– Wheel damping

7

ECS

www.esveld.com

Innorail2013

ENERGY DISTRIBUTED IN STRESS WAVES:

Rayleigh 67 % Shear 26 % Compression 7 %

GeometricalDamping Law

Vibrating source

Horiz.Comp.

RelativeAmplitude

Shear Wave

ν = 0.25

ShearWindow

GeometricalDamping Law

R-1

++

+ +

+

––

Compression wave

R-1

R-2 R-2 R-0.5

Rayleigh WaveVert.Comp.

8

ECS

www.esveld.com

Innorail2013

VIBRATION PROPAGATION AT GRADE

9

ECS

www.esveld.com

Innorail2013

VIBRATION PROPAGATION BY UNDERGROUND

10

ECS

www.esveld.com

Innorail2013

The human body is susceptible to the following frequencies: 0.1 - 0.2Hz : resonance of the organ of balance, resulting in

phenomena characteristic of seasickness; 4 - 8 Hz: resonance of the contents of abdomen and

thorax; 30 - 80 Hz: resonance of eyes in the eye sockets,

resulting in loss of focus;

The audibility limit lies at a frequency of approximately 20 Hz.

SUSCEPTIBILITY OF HUMAN BODY

11

ECS

www.esveld.com

Innorail2013CU

RVES

OF

EQU

ALVI

BRAT

ION

PERC

EPTI

ON

rms2

0.18 fEP v ( f )f1

5.6

= +

peak rmsv 2v=

2a 2 fv ( 2 f ) dπ π= =

102

101

8

5

3

2

8

5

3

2

100

8

5

3

2

10-1

101

100

10-1

5

3

2

8

5

3

2

8

5

3

2

8

Vibr

atio

n ve

loci

ty (r

ms-

valu

e) [m

m/s

]

Vibr

atio

n ve

loci

ty (p

eak

valu

e) [m

m/s

]

EP = equal perception

EP = 25.6

EP = 12.8

EP = 6.4

EP = 3.2

EP = 1.6

EP = 0.8EP = 0.56EP = 0.4

EP = 0.28

EP = 0.2

EP = 0.14

EP = 0.1

521 10 5020 100Frequency f [Hz]

Not noticeable

Very weak

Weak

Good

Strong

Very strong

12

ECS

www.esveld.com

Innorail2013

CLASSIFICATION OF VIBRATION LEVELS

ISOGERMANYReference v0not standardized LEP [dB] = 20 log10(v/v0), with for instance v0 = 10-8 m/s

13

ECS

www.esveld.com

Innorail2013

MEASURED VIBRATIONNEAR UNDERGROUND

14

ECS

www.esveld.com

Innorail2013

EFFECT OF HIGH-SPEED MEASURED AT DB

15

ECS

www.esveld.com

Innorail2013

mw cw kw F ( t )+ + =

( ) ( ) ( )π

≈

= =

1 kf2 m

K kw f F f H fTransmission Ratio

Mitigating measures:• Reduce force, cq excitation:

grinding, weld straightening;• Change natural frequency relative to

dominant excitation frequency:softer rail pads, resilient layers, …….

VIBRATION TRANSFER

k c

m

w

Fexcitation

Kresponse

10.00

1.00

0.10

0.01 0.1 1.0√2 10.0

f = excitation (impressed) frequency [Hz]f/fn

trans

mis

sion

ratio

|K(f)

/F(f)

|

ζ = 1.0

ζ = 0.5

ζ = 0.2

ζ = 0.1 ζ ζ = 0

16

ECS

www.esveld.com

Innorail2013

Dom

inan

t exc

itatio

n fre

quen

cy

Natu

ral f

requ

ency

stru

ctur

ef

f

f

1

→X

→=

Excitation spectrum

Transfer function

Response spectrum

VIBRATION TRANSFER II

excit naturalf 2 f>

17

ECS

www.esveld.com

Innorail2013

Low-frequency excitations are difficult to reduce

1 kf2 mπ

≈

Low stiffness and large mass necessary to achieve low frequency! Not with sleeper and rail pad, but with elastically supported slab.

PRACTICAL LIMITS

18

ECS

www.esveld.com

Innorail2013

Cork rubberConcrete block

Casting

BLOCK TRACK

19

ECS

www.esveld.com

Innorail2013

STEDEF

20

ECS

www.esveld.com

Innorail2013

PANDROL VANGUARD

21

ECS

www.esveld.com

Innorail2013

WHISPER RAIL THYSSENKRUPP

Vertical up to 10 mm Lateral < 2 mm

22

ECS

www.esveld.com

Innorail2013

KÖLNER EI

23

ECS

www.esveld.com

Innorail2013

ELASTIC RAIL SUPPORT

24

ECS

www.esveld.com

Innorail2013

SA 42 RAIL LOW NOISE -5 dB(A) compared to

ballasted track -7 dB(A) compared to

conventional slab track 60 % less consumption of

corkelast compared to UIC 54

25

ECS

www.esveld.com

Innorail2013

EMBEDDED RAIL ON STEEL BRIDGE

26

ECS

www.esveld.com

Innorail2013

SILENT BRIDGE WITH EMBEDDED RAIL

27

ECS

www.esveld.com

Innorail2013

SILENT BRIDGE VERSUS CONVENTIONAL

28

ECS

www.esveld.com

Innorail2013

CAST-IN SLEEPER

Ballast

Elastic supported sleeper

Polyurethane

Reinforced concrete

Filler concrete

29

ECS

www.esveld.com

Innorail2013

TRACK ON ELASTICALLY SUPPORTED SLAB

30

ECS

www.esveld.com

Innorail2013

GERB Floating Slab Track Systems

31

ECS

www.esveld.com

Innorail2013

Wege für MenschenUSP – Under Sleeper Pads

UNDER SLEEPER PAD (USP)

33

ECS

www.esveld.com

Innorail2013

UNDER SLEEPER PADS (USP) Increase of contact area between concrete and ballast (Riessberger): 5-10 % without USP ~ 35 % with USP

Adding resilience and thus reducing dynamic forces Very effective in areas with high impact forces: frog area of turnouts; transitions near engineering structures;

Effective solution at spots with maintenance problems due to poor subgrade;

37

ECS

www.esveld.com

Innorail2013

CONCLUSIONS Apply mitigating measures preferably at the source: smooth wheel-rail interface; sufficient track resilience;

Reduction of natural frequency: low spring stiffness; large mass;

1 kf2 mπ

≈