a unique field experiment to assess the noise ... - magleva unique field experiment to assess the...

UNIVERSITEITGENT

A unique field experiment to assess the noise annoyance

caused by maglev trains

Bert De Coensel, Luc Dekoninck,Tom De Muer, Dick Botteldooren

Acoustics Group, Ghent University, Belgium

Peter LercherDivision of Social Medicine, Medical University Innsbruck, Austria

Birgitta Berglund, Mats NilssonGösta Eckman Laboratory, Stockholm University, Sweden

Ministry of Transport, Public Works and Water ManagementDirectorate-General of Passenger Transport

Projectorganization ZuiderzeelijnDen Haag, Nederland

Ghent University – Medical University Innsbruck – Stockholm University 2

1. Introduction

2. Experiment

3. Results

4. Conclusion

Overview


Introduction

• Transportation noise annoyance- Trains ≠ highway traffic- Railway bonus for LAeq

• Valid for high-speed trains & maglev trains?- Vos (2004), Neugebauer et al. (1997), Fastl et al. (1996)

• Questions raised- Short samples used (45s) → temporal effect obscured- Small testperson panel → representativity- Nonacoustical factors (e.g. noise sensitivity) not taken into account

• The experiment presented- Conducted in a realistic setting (holiday cottage)- Exposure to longer fragments of sound + quiet periods- Traffic noise reproduced in ecologically valid way- Representative panellists selected using questionnaire


1. Introduction

2. Experiment

3. Results

4. Conclusion

Overview


Experiment

Field experiment in natural setting

Ecologically and socially valid setting

Good quality acoustic field reproduction

Panel of representative testpersons

• Natural setting- Holiday cottage in Westkapelle

(Zeeland, The Netherlands)- Quiet environment- Subgroups of panellists seated

in living room- Reading magazine, light

conversation, something to drink


Experiment





• Sound reproduction objectives- Realistic indoor 3D sound of

outdoor pass-by sources- Preserve natural feeling inside

experiment room


playback level

binaural recording B

Experiment





• Sound reproduction objectives• Methodology and validation

- Assumption: 2-channel recording- Checked for low speed trains at

short distance: 2 phases- Ideally: A = B

record level

binaural recording A


Experiment





• Sound reproduction objectives• Methodology and validation

- Assumption: 2-channel recording- Checked for low speed trains at

short distance: 2 phases- Ideally: A = B- Error within 5 dB in each

tertsband- Low frequency

05

101520253035404550

31.5 63 125 250 500 1k 2k 4k 8k 16kfrequency [Hz]

LAeq

,18s

ec [d

B(A)

]

realreproduced

Passage 1 - left ear

05

101520253035404550

31.5 63 125 250 500 1k 2k 4k 8k 16kfrequency [Hz]

LAeq

,23s

ec [d

B(A)

]

realreproduced

Passage 2 - left ear


Experiment





• Sound reproduction objectives• Methodology and validation• Reproduction setup

- 2 loudspeakers + subwoofer in front of slightly opened window

- Played on PC, equalized(31 bands) and amplified

- Façade level + indoor soundfield- No visual presentation


Experiment





• Sound reproduction objectives• Methodology and validation• Reproduction setup• Sample collection

distances: 25, 50, 100 and 200 m

mic

20 m

michead


Experiment





• Sound reproduction objectives• Methodology and validation• Reproduction setup• Sample collection

- IC: 140 km/h- TGV: 140 & 300 km/h- Maglev: 200, 300 & 400 km/h- Highway: free flow

• 45-sec passage fragments cut


Experiment





• Selection of panellists- Questionnaire to 1500 people in

neighbourhood of testhouse- Representative structure Dutch

population• RIVM environmental noise survey• Eurobarometer questionnaire

- Drawing 100 out of 255 replies• Age & hearing ability• Disimilarity using binary coding• Fuzzy resemblance Dutch person

0%

5%

10%

15%

20%

25%

0 1 2 3 4 5 6 7 8 9 10

noise sensitivity

participantsreference

0%

5%

10%

15%

20%

25%

0 1 2 3 4 5 6 7 8 9 10

quality of traffic noise

participantsreference


Experiment





• Listening test menus- Longer exposure → 10 minutes- Called “menu”- Consists of

• 2 passages of same train• 4 passages of same train• 8 minutes continuous highway noise

- Scaling context for panellists:• 7 reference 45-sec fragments• Produced of highway noise at 50 m• Scaled up and down + spectral


40

50

60

70

80

90

100

11:50 12:00 12:10 12:20 12:30Time [s]

Experiment





• Listening test menus• Outline test

- ± 5 panellists / session- Session façade level

• 14 minute training session• 7 menus of 10 minutes IC/TGV or

Maglev• 14 minute training session• 7 menus of 10 minutes Maglev or

IC/TGV• Conventional listening test

405060708090

100

10:10 10:20 10:30 10:40 10:50 11:00 11:10 11:20 11:30 11:40 11:50

Time [s]

405060708090

100

8:20 8:30 8:40 8:50 9:00 9:10 9:20 9:30 9:40 9:50 10:00Time [s]

LAeq

,1s

[dB

(A)]


Experiment





• Listening test menus• Outline test• Perceived noise annoyance

- Free number estimation- Master scaling

01020304050607080

40 50 60 70 80 90Road traffic reference sound level [dB(A)]

Anno

yanc

e

A = 70, assessment

R = 55, ms annoyance


1. Introduction

2. Experiment

3. Results

4. Conclusion

Overview


Results

• Average master scaled annoyance vs LAeq,10min

Sound level (LAeq,10min)

30 40 50 60 70 80 90

Annoya

nce

(m

aste

r sc

ale)

-10

0

10

20

30

40

50

60

70

Inter-cityMaglevRoad-trafficTGVMaster function

2 events / 10 minutes


Results



30 40 50 60 70 80 90

Annoya

nce

(m

aste

r sc

ale)

-10

0

10

20

30

40

50

60

70


4 events / 10 minutes


05

101520253035404550

55 60 65 70Sound level (LAeq,10min)

anno

yanc

e (m

aste

r sca

le) Conventional train

TGV high speedMaglev high speed

Results


• LAeq < 65 dB(A):- Maglev ~ IC

• LAeq > 60 dB(A):- Maglev ~ TGV

• 55 dB(A) → 65 dB(A):- Annoyance↓ [T,v,d]


Results

• Average master scaled annoyance vs LAeq,45sec

Sound level (LAeq,45)

30 40 50 60 70 80 90

Ann

oyan

ce (m

aste

r sca

le)

-10

0

10

20

30

40

50

60

70


• Assess loudness rather than annoyance

• Train passage at least as annoying as 45-sec highway

• LAeq > 60 dB(A):- Maglev + TGV hs

> TGV ls + IC

• LAeq > 65 dB(A):- Maglev ~ TGV


-10

0

10

20

30

40

50

60

70

80

30 40 50 60 70 80 90


Ann

oyan

ce (m

aste

r sca

le)

rise speed [ dB/s ]

Results


• Upward bending may be explained by rise time



30 40 50 60 70 80 90

Ann

noya

nce

(mas

ter s

cale

)

-10

0

10

20

30

40

50

60

70

80

Master function

Results


distance to track

• Annoyance less for trains passing at large distance


1. Introduction

2. Experiment

3. Results

4. Conclusion

Overview


Conclusion

• In “at home like” context, noise annoyance of ≠ types of trains at the same façade level is not significantly different

- LAeq < 65 dB(A): Maglev ~ IC- LAeq > 60 dB(A): Maglev ~ high speed TGV

• Railway bonus not observed for distances < 100 m

a unique field experiment to assess the noise ... - magleva unique field experiment to assess the...

Documents