Korrektion för vägyta vid beräkning av vägtrafikbuller

VTInotat

Nummer: TF 52-09 Datum: 1987-03-10

Titel: KORREKTION FÖR VÄGYTA VID BERÄKNING AV

VÄGTRAFIK-BULLER

Författare:

Ulf Sandberg

Avdelning: _

TF

Projektnummer:

52306-8

Projektnamn:

Reduktion av trafikbuller genom 'val av vägbeläggning

Uppdragsgivare:

Vägverket, Nordiska Bullergruppen, VTI

Distribution:

/ nyförvärv / begränsad / Nordiska Bullergruppen, VV

(db __ - _ _

' Statens väg- och trafikinstitut

- Pa: 58707 Linköping. Tel. 073-17 52 00. Te/ex 50725 VT/SG/ S _

Bakgrund

I den nordiska trafikbullermodellen tas ingen hänsyn till vägytans typ eller

beskaffenhet. Att sådan korrektion behövs, grunderna för vägytans

infly-tande på bullret samt vägytans karakterisering framgår av bilaga 1.

Statens Väg- och trafikinstitut (VTI) har gjort fordonsbullermätningar på

fordon i verklig trafik samt på olika vägytor, varefter mätningarna har

databearbetats för att få fram vägbeläggningsberoendet.

Denna PM redovisar hur en korrektion för vägytan bör utföras

kvantita-tivt. Som framgår av bilaga 1 är sambandet mellan trafikbullernivå och

vägyta en funktion av fordonstyp och hastighet. Samtidigt är de

erforder-liga korrektionerna inte försumbara varför metodiken med nödvändighet

blir komplicerad och således inte endast en tabell med en enda

korrek-tionsterm för varje vägtyp. Den korrektionsmatris som slutligen redovisas

kan kanske förenklas något genom sammanslagning av flera vägytetyper

men om och hur detta skall ske är snarast en "politisk" fråga och görs

därför inte här.

Mätmetod

På ett avstånd av 7,5 m från närmaste körfälts mitt och med mikrofonen

1,2 m över vägplanet har en ljudnivåmätare typ B 6: K 2209 med mikrofon

B 6: K 4165 varit monterad. Ljudet har även inspelats på band för senare

analys. Då enstaka fordon har passerat mikrofonen har maximala ljudnivån

avlästs samtidigt som fordonstyp (i åtta olika klasser) och hastighet har

registrerats. Genom att sammanställa dessa data kan statistik på

maxi-malnivån hos fordonen inom vissa fordonsklasser som funktion av

hastig-heten erhållas. Därigenom kan totala trafikbullret beräknas för godtycklig

trafiksammansättning och hastighet. Inte endast A-vägda ljudnivåer utan

även andra frekvensviktade värden kan tas fram, liksom frekvensspektra.

Tidkonstanten är 0,25 3 (motsv "Fast" på ljudnivåmätare). Mätningen

omfattar 100-150 personbilar och 50-100 tunga fordon per väg. Totalt har

ca 6000 fordonsbullermätningar gjorts.

Referensmaterial

Följande skrifter har utgjort referensmaterial:

l. SANDBERG, ULF; WENÄLL, JAN: The Influence on Traffic Noise of

Roads and Individual Vehicle Noise Emissions. Utförd åt Nordiska

Bullergruppen 1985-07-30, Statens väg- och trafikinstitut, Linköping.

2. VÄGVERKET: Samband mellan vägbeläggning och trafikbuller. PP

Meddelande Nr 26, Vägverket, Borlänge (1983). Även TR 6.492.02, IFM

Akustikbyrån AB .

3. ZETTERLING, TOMMY; AXEFORS, BENGT; WALTER, PER-OTTO:

Samband mellan torr och våt vägbana för trafikbuller. TR 7.638.01,

IFM Akustikbyrån AB, Stockholm (1984).

Utöver de mätningar som finns med i dessa har också beaktats mätningar

utförda 1986 och 1987 av VTI och kompletterande de mätningar som

redovisas i ref l, avseende:

Väg nr 18:

Tät, slät asfaltbetong, typ MAB16T, 8 år gammal, i

god kondition

Väg nr 47:

Enkel ytbehandling, Yi 12-20, 8 år gammal. Mycket

sliten.

.Väg nr 48:

Som väg nr 41, dvs dränerande asfaltbetong HABDlZ,

men 3 år gammal. Pga mycket intensiv trafik var

denna beläggning, trots ringa ålder, i senare delen av

sin livslängd.

Väg nr 49:

Bituminiserad chip-sten, BCS, med nervältad kvartsit

16-20 mm. Ålder: 9 år.

Väg nr 62

Dränerande asfaltbetong, HABDIZ; samma som nr 52,

Väg nr 63

53 och 54 i ref l, .fastän 2 år gamla. Utsatta för

Väg nr 64

måttlig trafik.

Korrektions matris

På nästa sida presenteras de korrektioner som är motiverade. Tabellen har

kompletterats undertill rörande vägytor som inte ingick i det aktuella

mätprogrammet men vars bulleregenskaper ändå i viss män är kända

genom däck/vägbanebuller mätningar.

Observera att korrektionerna utgår från en referensvägytetyp av samma

typ som bullermodellen är grundad på och inte från skillnaden mellan

uppmätt och predikterad ljudnivå. Egentligen borde "ref" i matrisen stå

för --1 dB om man tar hänsyn till nuvarande överskattning på ca 1 dB som

man har i modellens utgångsvärde (se VTI Notat TF 52-06).

Några vägytor har det bortsetts ifrån pga att de på senare tid har ersatts

av andra. Detta gäller främst vissa äldre dränerande typer, AEBlZÖ,

AEBlGÖ, HABlZÖ, HABléÖ. Dessa har numera vid nyanläggning i stort

sett utbytts mot HABDlZ.

I övriga nordiska länder förekommer en del vägbeläggningar som inte finns

med i korrektionsmatrisen och knappast förekommer i Sverige. Man får då

söka sig fram till den typ i matrisen som verkar vara mest besläktad med

den aktuella ytan. I ett fall är detta omöjligt, nämligen för

cementbetong-beläggningar vilka förekommer i Danmark. För en mycket slät och polerad

cementbetongbeläggning eller en tvärsräfflad beläggning torde

korrektio-nen vara ca +3 dB(A) över hela skalan. För en cementbetongväg med

hygglig finstruktur torde ca +1 dB(A) vara rimlig. Dessa syftade värden är

baserade på dels egna däckbullermätningar, dels belgiska och engelska

mätningar.

I Norge har gjorts försök med dränasfalt av något annorlunda typ än den

svenska HABD 12. Resultaten har visat sig lite mindre positiva än de

svenska, varför en korrektion enligt nr 4 bör användas även för en ny

beläggning.

Korrektionerna är egentligen uträknade för 0, 10 och 100 % tunga fordon.

För att undvika en ökad komplexitet har emellertid i matrisen en

generalisering till 0-5, 6-19 och 20-100 96 tunga fordon gjorts. I regel är

denna generalisering acceptabel. Emellertid blir "marginaleffekten" vid

t ex enkel ytbehandling (nr 7) vid ändring från 19 0/0 till 20 % tunga fordon

betydande (3 dB). I sådana fall bör en (logaritmisk) interpolation

4

Tabell 1

Korrektionstermer för vägyta i dB(A) vid fritt flytande trafik med

konstant hastighet.

Korrektionsterm

Vägbeläggningstyp

50 km/h

70 km/h

90-110 km/h

eller kondition

% tunga fordon

% tunga fordon

% tunga fordon

0-5 6-19 20-100

0-5 6-19 20-100

0-5 6-19 20-100

1.

Tät, slät asfaltbetong

ref

ref

ref

ref

ref

ref

_

ref

refref

minst ett år gammal,

t ex MABlZT, HABl6T

2.

D:o, men mindre än

0

-1 - -2

-2

-1

-1

-2

-2

-1

ett år gammal

3.

Dränerande asfaltbetong,

-4

-4

-4

-4

-4

-5

-5

-5

-5

HABDlZ, 0-1 år\

'

4. ' 13:0 medelålders

-2

_2

-2

-3

-3

-3

-3

-3 -3

5.

13:0, gammal

0

-1

-2

0

-1

-2

-1

-2

-2

6.

Dubbel ytbehandling, v2,

+2

0

.-1

+1

0 > -2

0

0 ..2

översta lagret 4-8 mm,

skrovlig

7.

Enkel ytbehandling,

+3

+1

-2

+3

+1

-1

+2

+1

0

Yl 12-16 eller

12-20 mm, 0-1 år

8.

D:o, sliten

+2

+l

-1

+2

+1

0

+2

+1

0

9.

Bituminiserad Chip-sten,

+2

+1 .

0

+2

+1

0

+2

+1

0

BCS, minst 1 år gammal

10.

Som nr 1, men våt

+6

+5

+3

+4

+4

+2

+3

+2

+2

11.

Som nr 3, men våt

0

-l

-2

-l

-2

-2

-2

-2

-2

11.'

Som nr 1, men vinter-

+4

, +3

0

+4

+3

+1

+3

+3

+1

förhållanden med minst 50 % dubbdäck

13.

Somnr 3, men vinter-

0

-1

-2

0

-1

-2

-1

-1

-l

förhållanden med minst

50 96 dubbdäck

Övriga, indirekt uppmätta vägytor:

Enkel ytbehandling, Yl, med 8-12 mm sten torde motsvara ungefär nr 6 ovan.

Nr 6 i slitet tillstånd torde motsvaras närmast av nr 8.

Oljegrus, o.d. torde motsvaras av nr 7.

Dränasfalt av typ Drainor 12 torde mest likna nr 3-5.

Rubit-beläggningar torde mest motsvaras av nr 1-2.

Våt-vägs-korrektionen, nr 10, kan troligen gälla för samtliga vägytor utom nr

3-5 (de dränerande). Den är dock starkt avhängig regnintensiteten (värdena här

gäller ett kraftigt regnväder).

Vinter-korrektionen, nr 12, kan likaså troligen gälla för samtliga vägytor utom

nr 3-5.

Behovet av vägtexturmätning

För att få en bättre noggrannhet i bullerpredikteringen borde egentligen

den aktuella vägytans textur vara känd (eller predikterad). Emellertid är

korrektionerna mer än nog komplicerade redan utan en sådan

texturhän-syn. Den ökade noggrannheten av en vägtexturspecifikation torde

knap-past stå i rimlig proportion till det ökade arbetet och till andra

onog-grannheter samt till risken för misstag, varför tanken på en

vägtexturspe-cifikation har släppts tills vidare.

Ovrigt

' Avvikelser gentemot ref. nr 2 förekommer. Det gäller följande:

a)

Vissa dränerande beläggningar som ingår i ref 2 var experimentella

och är ej längre aktuella eftersom bättre alternativ har erhållits.

Resultaten beaktas därför ej här.

b)

Ref 2 har en mindre korrektion än 1 här presenterad tabell för

personbilar vid 50 km/h för "grov ytbehandling". VTIs resultat bör

vara mer tillförlitliga av den anledningen att de är differentierade

på lätta och tunga fordon och troligen omfattar flera fordon.

c)

För dränasfalt har Ref 2 alltför låga korrektioner, troligen beroende

på att en del äldre, experimentella vägytor ingår eller att ytorna

inte var nya vid mätningen.. Se fö kommentar ovan ang norsk

dränasfalt.

Rekommendation

En korrektion för vägyta (samt våtvägs- och dubbdäcksförhållanden) bör

införas i den nordiska trafikbullermodellen vad avser fritt flytande trafik

50-110 km/h. För accelerande trafik 0-50 km/h kan än så länge ingen

korrektion anges. Den blir under alla omständigheter avsevärt mindre än

för'fritt flytande trafik och kan tills vidare försummas. På lång sikt kan

emellertid en korrektiOn behövas även där, vilket har antytts av vissa

provningar enligt ISO 362 och ISO 7188.

Exakt hur korrektionen skall införas bör diskuteras utgående från här

presenterade data. Den frågan är av praktisk och "politisk" natur och

berörs ej här. Man bör t ex beakta datoriseringsönskemålen. I en

datorise-rad modell kan man t ex mata in termerna för 0 och 100 % tunga fordon

och låta datorn göra alla interpolationer. Sedan man bestämt hur

korrek-tionen skall göras kan kanske en något mer praktisk (?) vägytebeskrivning

utformas som tar hänsyn till den valda utformningen.

Man kan mycket väl tänka sig att en sida i dokumentet som beskriver

modellen utformas ungefär som korrektionsmatrisen men med notiserna

under densamma inarbetade inom ramen i texten under

"vägbeläggnings-typ". Så har inte gjorts här, för att markera att data är ofullständiga.

Dessutom bör i resp språkversion vissa danska och norska motsvarigheter

beaktas.

W

Applied Acoustics 21 (1987) 000«000

Road Traffic Noise-The Influence of the Road Surface

and its Characterization

Ulf Sandberg

Swedish Road and Traffic Research Institute, 8-58101 Linkoeping (Sweden)

SUMMARY

Unacceptable errors in theprediction of trañc noise occur in some cases when

the road surface is large!y dWerentfrom that on which theprediction model is

based. The reason is that tyre/road noise has appeared to be the dominating

component of the noise from free-flowing traffic and that this noise is to a

substantial extent dependent on the road surface.

The mechanisms for ter/road noise. generation and its relation to road

Characteristics are described. Relevant road surface Characterization

methods are suggested. The major method is the measurement of the road

texture profle and subsequent spectral analysis of the prof/e curve.

Supplem'entary methods concern the measurement of acoustical and

mechanical impedances. It is concluded that the road surface efect on trañc

noise is extremely complicated and that it is very diâicult to generalize any

simple relations.

.

,

For free-filowing trañc it is shown that the tested road surface types and

conditions may influence the traâic noise by up to 11dB(A ). This callsfor a

correction term for the road surface in the prediction models. Despite the

complicated relations, it appears feasible-within stringent limitations--to

use a table where the correction term is a variable ofvehicle type, vehicle speed

as well as road surface typeand condition.

'

1 INTRODUCTION

Some of the procedures used for the prediction of traffic noise are so

accurate today that often they may give a noise level more reliable and

representative than the noise level obtained when actually measuring at the

speciñc site. However, in some cases large errors in theprediction may occur.

Such a case is when the tyre/road noise is atypical of the condition assumed .

'

l

Applied Acoustics 0003-682X/87/SO3-50 © Elsevier Applied Science Publishers Ltd, England,

1987. Printed in Great Britain

'

i

2 Ul fS an db er g NO . OF HE AS UR ED RO AD S i; ME AS UR EO -_ . PR ED IC TE D ' ' : NO CS E L E V E L ' '. . _ I. 0 -. I \ 1 l I I ' I '6 '4 "Z 0 *2 dBl Å)

/ \ \ / DR AI NA GE SU RF AC ES o m mSU RF AC ES Fi g. 1. Th e di ff er en ce be twe en me as ured an d pr ed ic te d tr af fi c no is e le ve ls (e qui va le nt A-we ig ht ed le ve ls ). Th e pr ed ic ti on mo de l is th e No rd ic mo de l4 an d th e da ta we re me as ur ed by th e aut ho r be twe en 19 83 an d 19 85 . /\ V n or ma l in th ep re di ct io n mo de l, fo ri ns ta nc e wh en sp ec ia l ro ad sur fa ce s ar e us ed (s ee Fi g. 1). Th is im pl ie s th at tyr e/ ro ad no is e is an im po rt an t co nt ri but or to th e ove ra ll tr af fi c no is e. As tyr e/ ro ad no is e is la rg el y in ñue nc ed by ro ad sur fa ce Ch ar ac te ri st ic s, an ot he r im pl ic at io n is th at pr ed ic ti on mo de ls sh oul d in cl ude a co rr ec ti on te rmfo r ro ad sur fa ce in fl ue nc e. Se ve ra l mo de ls in de ed ha ve th is po ss ib il it y, fo ri ns ta nc et he Br it is hp ro ce dur eCa lc ul at io n of Ro ad Tr af fi cN oi se ,l wh er e a co rr ec t' m of up to +4 dB(A ) is re qui re d fo r de ep ly gr oo ve d ce me nt co nc re te sur fa ce s, an d the Ne th er la nd 's Ro ad Tr af fi c No is e Ca lc ul at io n Pr oc ed ur e, 2 wh er e co rr ec tion s up to +4 '5 dB (A ) ar e re qui re d. In or de r to -en ab le co rr ec ti on s to th e mo de l, th e ro ad sur fa ce mus t be cl as si ñed or qua nt iñe d in so me wa y. Ge ne ra ll y, in th e mo de ls kn own to th e aut ho r, on lya fe w ext re me typ es of sur fa ce s or pa ve me nt s ar e de ñn ed but in on e ca se th em ea sur em en t of th e ro ad text ur e is al so sp ec iñe d. Th e la tt er is in th e Aus tr al ia n St an da rd , wh er e th e sa nd -p at ch me th od fo r te xt ur e me as ur em en t is' re co mm en de d to Sup pl em en t th e ro ad typ e cl as si ñc at io n. 3 Th is pa pe r 'r ep or ts me as ur em en ts an d ca lc ul at io ns of im po rt an ce fo r es ti ma ti ng th e ro adsur fa ce in fl ue nc e on tr af fi cn oi se as we ll as di sc us si on of th e co mp le xi ty of th is sub je ct . 2 C O N T R I B U T I O N O F T Y R E / R O A D N O I S E T O T R A F F I C N O I S E Tr af fi c no is e m a y ha ve two ba si c or ig in s: th e po we r-tr ai n un it of th e ve hi cl e (e ng in e, ai r in ta ke , exh aus t, tr an sm is si on ) a n d th e tyr e/ ro ad in te ra ct io n. T h e Th e fr i/ fa me :) of th e road Sur fa ce an d it s Ch ar ac te ri za ti on 3 co nt ri but io n of tyr e/ road no is e or po we r-tr ai nno is e to to ta l no is e fr om a traf fi cs tr ea m ca n on lybe ca lc ul at ed , no t me asur ed . In a st ud y ma dein 19 81 by the aut ho r, th e No rd ic tr af fic no is e pr ed ic ti on mo de lwa s us ed to pr ed ic t

th

e

to

ta

l

tra

ffi

cn

oi

se

in

fo

ur

si

tua

ti

on

sz'

*'

5

l. Ave ra gesp ee d 70 km /h , t yp ic al'o fb y-pa ss ro ad s in and ar oun d ci ti es or high -c ap ac it y ro ad s ne ar sub ur ba n re si de nt ia l ar ea s. 2. Ave ra ge sp ee d 50 km /h ,t yp ic al of no n-re st ri ct ed ur ba n tr af fi c. Ave ra ge sp ee d 30 km /h ,t yp ic al of co ng es te d urba n tr af fi c. 4. Ave ra ge sp ee d <3 O km /h ,vi ci ni ty of a li gh t-cont ro ll ed cr os si ng wh er e ve hi cl esar e ei th er de ce le ra ti ng or ac ce le ra ti ng . rr; Ea rl ie r an d la ter me as ur em en ts have pr ove n th at th isNo rd ic pr ed ic ti on me th od isve ry ac cur at e in th isca se .6 Tr af fi c no is e data we re gi ve n in c h (d B( A) ).Th en th e no is e ca used by tyr e/ ro ad in tera ct io n fo rt he same tr af fi c situa ti on s wa s ca lc ul ated , fr om pe ak by-pass le ve ls to c h le vels , us in g me as ur ed da ta of in di vidua l ve hi cl es .T he road sur fa ce wa s as sum ed to be a sm oo th as ph al t conc re te typ ic al of urba n ro ad s. Th re e cond it io ns we re co ns id er ed : d ry road , 'we t ro ad ' an d stud se as on . Th e la st conc er ns th e us e of st ud de d tyres on 70 % of th e cars as typ ic al of Swe den. Swe di shwe at he r an d st ud de dtyr e st at is ti cs we re used to ca lc ul at e th e ye ar ly tyr e/ro ad no is e, le ve l ba sed on th e en er gy pr inci pl e, th e s0 c al le d ye ar ly eq ui vale nt le ve l (Y EL ). Wh en tyr e/ ro adno is e ch ha d be en ca lc ul at ed it wa s comp ar ed to th e ove ra ll tr af hc nois e Le a. Th e di ff er en ce was us ed to ca lc ul at e po we r-tr ai n no is e. Th e re sul ts ar e sum mari ze d in Fi g. 2. It ca n be se en th at th e ye ar ly LM 1 of tyr e/ ro ad no is e (Y EL ) domi na te s ove r po we r-tr ai n no is e fo r bo th 50 an d l Le o HV -5: lt HV -10 1 i HV a 20 1 65* ldB tAl l _. est ss_ , 60 -60 -55 * 55 ..

/ TR AF F 1C 7 .' s n um o x <3 0 30 50 70 [m m] (3 0 30 s'o 7'0 _-- -PO WE R-TR Ä! " NO IS E TY RE/R OA D NO IS E: YE L (Y EA RL Y Le o) -' T Y R E / R O A D NO lS E: DR Y RO AD -TY RE /R OA DNO lS E: ST UD SE AS ON .n u -TY RE /R OA DNO IS E: HE T RO ÅD Fig. 1 Th e co nt ribut io n to tr af fi cno is e fr om po we r-tr ai n an d tyr e/ ro ad so ur ce s. fo r di ff er en t pc rc cntug cs of he avy ve hi cl es (H V) and in fo ur di ff eren t tr af fi c si tua tion s.

4 Ul f Sa nd be rg 70 km /h ! It is no t ne glig ib le eve n at 30 km /h . Fo r warm er an d dr ie r cl im at es , th e dr y ro ad case mi gh t be ap pr oa ch ed , but eve n th en tyr e/ ro ad no is eis ap pr oxi ma te ly eq ua lto po we r-tr ai n no is e at 50 km /h. ln co un tr ie s wh er e th e dr ivi ng be ha vi our is sub st an ti al ly di ff er en t, th ere sul t co ul d of co ur se be di ff er ent. Inla te r ye ar s, th es e me as ur em ents an d ca lc ul at io ns ha ve be en ch ec ke d by th e aut ho r ag ai ns t re ce nt me as ur emen ts wh ic h ha ve co nñr me dth e da ta gi ve n ab ove .6 Toge th er wi th th e kn owl ed ge th at ro ad sur fa ce s in fl ue nc e tr af fi c no is eas muc h as tyr es , th e ab ove de mo ns tr at es th e im po rt ance of co ns id er in g ro ad sur face s wh en pr ed ic ti ng an dab at in g tr af fi c no is e. 3 C ÖR R E L A T I ON B E T W E E N T Y R E / R O A D NO IS E A N DR O A D T E X T U R E 3. 1 No is e an d te xt ur e: sp ec tr al co rr 'e la ti ons So me ye ar s ag o, th e Swe dish Ro ad an d Tr af ii c Re se ar ch In st it ut e an dth e Be lg ia n Ro ad Re se ar ch Ce nt er appe ar ed to ha ve pe rf or me d two al mo st id en ti ca l re se ar ch pr og ra mm es on th e re la ti on be twe en no is e an d ro ad Ch ar ac te ri st ic s. The re sul ts tur ne d out to be ful ly co mp at ib lean d a po ol in g of th e da ta as we ll asla te rs up pl em en ta ry st ud ie dyi el de d th e re sul ts re po rt ed he re (see al so re fs 7 an d 8). s Tyr e/ ro ad no is e m a xi m um le ve ls at co as t-by (7 -5 m fr om th e mi cr op ho ne ) we reme as ur ed fo rf our pa ss en ge r-ca r tyr e typ es onse ve ra l ro ad pa ve me nt s at two or mo re sp ee ds . Th e ro ad sur fa ces We re ch ar ac te ri ze d by me as ur ing th eir lo ng it ud in al pr of il e (s pe ct ra l an al ys is in 1/ 3 or .l /1 2 oc ta ve ba nd s) , fr ic ti on coeñi ci en t, wa te r out -f lo w ti me (a dr ai na ge meas ur em en t) , sa nd -pa tc h text ur e de pt h (o nl y Be lg ia n da ta ) an d soun d ab so rp ti on or pr 0p ag at io n. Th e mo st ob vious co rr el at io n wa s fo un d to be th at be twe en th e tyr e/ ro ad no is e sp ec tr al le ve ls an d th e ro ad te xt ur e pr of il e sp ec tr alle ve ls . He re ,s oun d pr es sur e le ve ls we re gi ve n in 25 th ir d-oc ta ve ba nd s wi th ce nt er freq ue nc y f Ro ad te xt ur e leve ls we re al so gi ve n in 25 th ir d-oc ta ve ba nd s, wi th ce nt er wa ve le ng th 2.. Fo ra ce rt ai n ac ous ti c fr eq ue nc y f an d a ce rt ai n te xt ur e wa ve le ng th i., we ha ve n pa ir s of so un dpr es sur e le ve ls (S PL ) an d ro ad te xt ur e le ve ls (T L) (wh er e n = num be r of ro ad sur fa ce s) . _ ' A li ne ar re gr es si on anal ys is wa s ma de on th es e da ta pa ir s (S PL ; TL ) fo r th e di ff er en t pa ve me nt s. Th e co rr el at io n co eñâc ie nt (r ) betwe en S P L an d T L va lue s wa s ca lc ulat ed . Th is ca lc ul at io n wa s re pe at ed fo r di lT eren t (S PL ; T L ) 77 20in fh wn ce Q/ 'I hc m ad sur fa ce an d it s Char ac te ri za ti on 5 " R E X " r t " m um T l R E w n u" . so : Im ' (H 11 31 50 . 10 00 . , ' 5 4 10 0. M A .5 \ :i s .6 0 32 to 1( mr n)32 32 5 16 0 52 ioN m ) :3 *se ..5 Fi g.3. Co nt our li ne sof th e co rr el at io nco eñ' ic ie nt be twe en no is e le ve l fo r four tyr es (8 0 km / h) and te xt ur e le ve l. Circ le d num be rs deno te ext re me va lues at 'c ri ti ca l' ac ousti c fr eq ue nc ie s an d te xtur e wa ve le ng ths. Tyr eX =t yr ewi th 's um me rtr ea d' ; T yr e W =tyr e wi th 'wi nter tr ea d' ; Tyr e S = tyr e wi ths um me r tr ead' ; Tyr e P = smoo th , no n-pa tter ne d tyr e. data se ts ,e ac hse t co rr es pond in g to ace rt ai nf and i. co mb in atio n. W e were then ab le to list th e two -va ña bl e corr el at io n coef ñc ie nt

r=

RÅÅ)

(1)

As we ha ve 25 th ir d-oc tave ba nd sof ac ous ti cfr eq ue nc yan d te xt ure wa ve leng th , re spec ti ve ly, th e fi na lre sul t wi llbe a 25x 25 ma trix= 62 5 va lue sfo re ac h spee d an d each tyr e. Figur e 3 sh ows a pl ot ofth is re la tion fo r 'f our tyres an d one sp ee d (8 0km /h ) in the fo rm of fve rs us Å for so me co nsta nt value s of r (con to ur li nes) . In te rp olat io n ha sbe en us ed wh er e da taha ve to o po or re so lut io n. Th e re sults ar e im port an t fo rth e di sc ussi on fo ll owi ng an d can be sum ma rize d as fo ll ows : 1. Th e ge ne ral pa tt er n isqui te si mi lar fo r al l th efo ur tyr es . 2. Th e co rrel at io n ca nbe bo th posi ti ve (n oi sein cr ea se s with te xt ur e) an d ne gati ve -( no ise de cr ea ses wi th text ur e) . 3. Th e freq ue nc ywh er e the co rr el atio n ch ange s si gn is ca ll ed th e cros so ve rfr eq ue ncy an d is ap pr oxi ma te ly 10 00 -1 250 Hz fo r pa ss enge r-ca rtyr es .

6 Ul fS an db er g 4. Th e co rr el at io n is ma xi ma ll y po si ti ve be twe en no is e an d te xt ur e at 40 0-50 0 Hz an d 63 -8 0 m m te xt ur e wa ve le ng th , re sp ec ti ve ly. 5. Th e co rr el at io n is ma xi ma ll y ne ga ti ve be twe en no is e an d te xt ur e at 25 00 -8 00 0 Hz an d 2-32 mm , re sp ec ti ve ly. Th e co nc lus io n is th at th e no is e is ge ne ra te d by so me pr oc es s cl os el y co nn ec te d wi th th e ro ad te xt ur e suc h th at at lo w fr eq ue nc ie s th e no is e is pr op or ti on al to te xt ur e, a re la ti on wh ic h is in ve rs e at hi gh fr eq ue nc ie s. 3. 2 Th e cr it ic al no is e fr eq ue nc ie s an d te xt ur e wa ve le ng th s Th e ac ous ti c fr eq ue nc ie s an d te xt ur e wa ve le ng th s at wh ic h th e co rr el at io n ha s ext re me va lue s we re ca ll ed cr it ic al fr eq ue nc ie s (f *) an d cr it ic al l i-ra ve le ng th s( ).* ). As th er e we re ext re me va lue s bo th in th e lo w (L F) an d hi gh (H F) fr eq ue nc y ra ng es , th e co rr es po nd in g cr it ic al fr eq ue nc ie s an d wa ve le ng th s we re ca ll ed L an d H, re sp ec ti ve ly. If th e L F an d H F no is e-ge ne ra ti on me ch an is ms ar e re la te d, th er e sh oul d be a co rr el at io n be twe en th e LF an d H F no is e le ve ls , pa rt ic ul ar ly at th e cr it ic al lo w an d hi gh fr eq ue nc ie s, fL * re sp ec ti ve ly få. Ho we ve r, fo r a gi ve n tyr e an d sp ee d, th e SP Ls at th e cr it ic al fr eq ue nc ie s we re no t fo un d to be co rr el at ed (T ab le 1) ; th is co nf ir ms th at tyr e/ ro ad no is e is th e sup er po si ti on of two in de pe nd en tl y ge ne ra te d sp ec tr a. Th us th e L F an d H F no is e-ge ne ra ti on me ch an is ms ar e no t co rr el at ed . In fa ct , th ey ha ve co nñi ct in g re la ti on s wi th ro ad te xt ur e. It co ul d be pr ed ic te d th at ro ad te xt ur e wi th a wa ve le ng th i. sh oul d ge ne ra te a no is e wi th a fr eq ue nc y f if th e tyr e is ro ll in g wi th a sp ee d v ac co rd in g to th e re la ti on

f=

12/

1

'

(2)

if fh as th e un it Hz, v' th e un it m/ s an d 2. th e un it m. Th er e is a te nd en cy in Fi g. 3 th at th e co rr el at io n is ve ry hi gh al on g li ne s T A B L E 1 Co rr el at io n Co eñi ci en ts (r) Be twe en th e So un d Pr es sur e Le ve ls at th e Cr it ic al Fr eq ue nc ie s I ? an d. Ä! (N ot e th at no co rr el at io n is si gn if ic an t eve n at th e 1 0 % ri sk le ve l. T h e fo ur tyr e typ es (P ,S ,W ,X ) ar e de sc ri be d in th e ca pt io n to Fi g. 3.) Pa ra me te r 7) 'r e typ e P 5 W X X X X X Q Sp ee d (k m/ h) 80 80 80 40 60 80 10 0 12 0 r -0 -1 8 -0 41 -0 2 9 -O -l O -0 -1 0 -0 0 7 -0 0 6 -0 -0 5 Sa mp le si ze 16 16 16 33 33 33 33 33 Th e in fl ue nc e of th e ro ad sur fa ce an d it sC hara ct er iza ti on 7 A c o us t i c fr eq ue nc y (f ) [H 2] 1 0 k 1 Sk -t 2k 4 50 0 -1 0 0 1 (Å) Te xt ur e wa ve le ng th I 1' I T 1 T I 20 0 10 0 50 20 '1 0 5 2 [m m] Fi g. 4. The re la ti on be twe en ac ous ti c fr eq ue nc y an d te xt ur e wa ve le ng th fo r wh ic h th er e is ma xi mum co rr el at io n be twe en no is e an d te xt ur e le ve ls . Fr eq ue nc ie s an d wa ve le ng ths ar e in th ir d-oc ta ve ba nd s. Th e cur ve re pr es en ts th e ave ra ge fo r th e fo ur te st ed typ es . de sc ri bi ng th e ab ove re la ti on betwe en f an d 2.. Fo r exa mp le , fo r tyr e 8, r al on g th e li nef = U/ Å 'is wi th in 10 % of th e ma xi mum fr om at le as t 10 0-80 0Hz. Fo r all fo ur tyr es , th e ma xi mum al wa ys li es wi th in one oc ta ve fr om th at li ne fo r al l fr eq ue nc ie sup to an d in cl ud in g 80 01 -1 2. Ho we ve r, fo r al l tyr es , an d fo r fr eq ue nc ie s ab ove 12 50 -2 00 0 Hz, th e m a xi m um do es no t fo ll ow th e li ne f = v/i .but is co ns ta nt at wa ve le ng th 2-32 mm . Fo r a gi ve n ac ous ti c fr eq ue nc y, th e te xt ure wa ve le ng th at wh ic h th er e is ma xi mum co rr el at io n be twe en no is e and te xt ur e le ve ls fo ll ows th e pa tt er n sh own in Fi g. 4. A st ud y of th e m a xi m um co rr el at io n as a fun ct io nof Sp ee d ve ri ñe s th at th e fr eq ue nc ie s fo rm a xi m um co rr el at io n ar e sh if te d ap pr oxi ma te ly pr op or ti on al to th e sp ee d in th e LF rang e. Al l th is sup po rt s th e hyp ot he si s th at in th e L F ra nge (o nl y) th er e is ,a re la ti on f = v/ Å, i.e . th e acous ti c fr eq ue nc y eq ua ls th e fr eq ue nc y of te xt ur e im pa ct on a ro ll in g tyr e. ' Th e L F me ch anis m th en is th e im pa ct of road ch ip pi ng s on th e tyr e tr ead or Vi ce -ve rs a, i.e .t he im pa ct of tyr e trea d el em en ts on th e ro ad sur fa ce ,wh ic h in duc es ra di al vi br at io ns in th e tyr e.Th e H F me ch an is m isin de pe nd en t 0....--" _.r. i

8 Ul f Sa nd be rg fr om th e LF me ch an is m, but co rr el at ed wi th ro ad te xt ur e at sh or t te xt ur e wa ve le ng th s an d is be li eve d to be so me ai r-pum pi ng or ai r-re so na nt me ch an is m pe rh ap s exc it ed by st ic k-Sl ip mo ti on s in th e tyr e/ ro ad in te rf ac e (s ee ref .7 ). It is po ss ib le th at th e H F me ch an is m exp la in ed he re ma y in fa ct be th e sum of se ve ra l me ch an is ms , but th e re la ti on wi th th e ro ad sur fa ce se em s to be sub st an ti al ly un am bi guo us . Our re sul ts we re ob ta in ed wi th pa ss en ge r-ca r tyr es . La te r, re se ar ch er s in th eU S9 ap pl ie d our te ch ni que to tr uc k tyr es an d we re ab le to ve ri fy th at our re sul ts we re va li d al so fo r tr uc k tyr es . Ho we ve r, it se em s li ke th e cr it ic al fr eq ue nc ie s, cr it ic al te xt ur e wa ve le ng th s an d th e cr os so ve r fr eq ue nc ie s ar e all sh if te da fa ct or of 2, wh ic h is th es am e fa ct or as be twe en th ep hys ic al si ze s of tr uc k an d ca r tyr es . 3. 3 Ot he r in fl ue nc es of th e ro ad sur fa ce In ad di ti on to th e me ch an is ms de sc ri be d ab ove , th e fo ll owi ng am pl it ud e in fl ue nc in g ef fe ct s ha ve be en id en ti fi ed : 1. Th e ho rn ef fe ct . 2. So un d ab so rp ti on in th e ro ad sur fa ce . 3. Me ch an ic al im pe da nc e ef fe ct . Be twe en th e cur ve d tyr e tr ea d fo re an d af t of th e tyr e/ ro ad in te rf ac e an d th e ro ad sur fa ce th er e is a sp ac e fo rm in g an ac ous ti ca l ho rn wh ic h in cr ea se s th e ra di at io n ef fi ci en cy ba ck wa rd s an d fo rwa rd s. lo Th is ho rn ef fe ct mi gh t be ef fe ct ive r el im in at ed wh en on e of th e sur fa ce s of th e ho rn is po ro us suc h as wh en th e ro ad pa ve me nt is a so -c al le d dr ai na ge as ph al t. Th e so un d ab so rp ti on ef fe ct 7' 14 oc cur s on ly on so -c al le dd ra in ag e as ph al t (p or ous or pe rvi ous as ph al t ar e al te rn at ive na me s) wh er e th e sur fa ce ha s a si gn if ic an t so un d ab so rp ti on . It in ñue nc es no t on lyt yr e/ ro ad no is e but al so po we r-tr ai n no is e. ll Th e so un d ab so rp ti on ef fe ct is of gr ea t im po rt an ce to tr af fi c no is e. -. _v A hyp ot he si s wh ic h is sti ll no t re je ct ed or suf fi ci en tl y ve ri fi ed is th at th e st if fn es so f th e ro ad sur fa ce , or th e ma tc hi ng of me ch an ic al im pe da nc e tyr e-to -r oa d, in ñue nc es th e tr ea d bl oc k or ro ad te xt ur e im pa ct so th at it mi gh t be am pl iñe d (st iff ro ad ) or at te nua te d (s of t ro ad ). 12 It im pl ie s th at ri gi d pa ve me nt s li ke ce me nt co nc re te ma y be so me wh at no is ie r th an ñe xi bl e pa ve me nt s li ke as ph al t co nc re te , an d th at th e no is e m a y in cr ea se so me wh at wh en a sur fa ce is ag ed by co mp ac ti on . I3 .4 Im pl ic at io ns fo r th e ove ra ll A-we ig ht ed no is e le ve ls W h e n co ns id er in g_ th e ab oV e-'m en ti on ed c on fl ic ti ng ' re la ti on s be twe en hi gh an d lo w fr eq ue nc ie s of tyr ef ra ad no is e an d lo ng an d sh or t wa ve le ng th s of te xt ur e, it is ob vi ous th at th e in ñue nc e of te xt ur e on th e ove ra ll A-we ig ht ed Th e in fl ue nc e of th e road sur f/ ac e an d it sC ha ra ct er iza ti on 9 no is e is de pe nd ent on th e re la ti ve im po rt ance of th e hi gh -an d lo w-fr eq ue nc y no ns e co mp on ents . Al so , th e di ff er en ces ob se rve d be twe en ca ran d tr uc k tyres co nc er ni ng cr os so ver fr eq ue nc y co mp li cate th e si tua ti on . Itis no t po ss ib le to ge ne rali ze an d sa y th at a ro ughe r te xt ur e me an s a hi gh er A-we ig ht ed nois e le ve l, Or vi ce -ve rs a. Suc h ge ne ra li za ti on sar e poss ib le on ly wi th in ce rt ain li mi ts an d fo r ce rt ai n text ur es , an d th en it is no long er a ge ne ra li za ti on . Fo r in st an ce , wh en co nsid er in g on ly ve ry ro ug h te xt ur es , th e A-we ig ht ed no is e le ve l is do mi na ted by th e lo w-fr eq ue nc y co mp on en ts . In th is ca se on e co ul d sa y th at , with in a na rr ow ra ng e of te xt ur e, Ao we ig ht ed le ve ls in cr ea se wi th te xt ure. Th e co nt ra ry is tr ue for ve ry sm oo th te xt ur es , beca us e th e ove ra ll le ve l isth en do mi na te d by th ehi gh -fr eq ue nc yco mp on en ts . Wh en cl as sifyi ng tyr es fr om th e poin t of vi ew of no is e,th e Cl as si fi ca ti on m a ybe co mp le te ly di ff er en t,eve n re ve rs ed , on di ff er ent ro ad sur fa ce s. 13 A si mi la r eve nt ma y occur fo r Cl as si fi ca ti on of road s. A ro ad wh ic h is si lent fo r tr uc ks m a yb e nois y fo r ca rs , as il lustr at ed in Fi g. 12 . Ap ar t fr om th eab ove fa ct or s, sp ee dal so in fi ue nc es th e re lati ve im po rt an ce of th elo w-an d hi gh -f re que nc yme ch an is ms .7 Fi na ll y, thee ff ec ts of we t ro ad s.th e ef fe ct of di ff er en td ir ecti on al Ch ar ac te ri st ic so fthe no is ea nd th e ef fe ct of ch ange d sp ec tr al Ch ar ac te ri stic s wh en th e no is e is tr ansm it te d thro ug h a fa ça de ha ve al lno t be en me nt io ne d. The la tt er m a y me an th at a ro ad sur fa ce wh ic h is n oi sy out si de a ho us e mi gh tbe re la ti ve ly *s il en t* in si de it, an d vi ce -ve rs a. 4 In co nc lusio n, if th ea im is topr es en t si mp le re la ti onsb et we en tr af fi cn oi se an d road sur fa ce si t is ob vi ous th at th e mo re one le ar ns ab out tyr e/ road no is e the mo re on e ap pr oa ch esth e st at e of ch ao s. Th e safe st ge ne ra li za ti on wo ul dbe to sa y th at th e foll owi ng re la ti on se ems to be tr ue fo r tyr e/ro ad no is e: Si mp li ci ty an d ap pr eh ensi ve ne ss ar e ei th er th e li mi ta ti on s. of pr es en te dre sul ts an d th ei r prop or ti on al of th e da ta, or th e im pl ic at io ns -to ig no ra nc eof th e aut ho r 4 R O A D S U R F A C EC H A R A C T E R I Z A T I O N 4. 1 Ge ne ra l Fo r a co mp lete ro ad sur fa ceCh ar ac te ri za ti on with re sp ec t to no ise on e sh oul d, as a co ns eq uenc e of th e pr evi ous se ct io n, me as ur e the fo ll owi ng qua ntit ie s: 1. Th e te xt ur e prof il e sp ec tr um . 2. Th e so un d abso rp ti on co ef fi ci en tor so un d pr op ag at ion. 3. Th e me chan ic al st if fnes s or im pe danc e. ...-...øu v . ._ .-.-...n-.-_ _._. ._

10 Ul f Sa nd be rg 4. 2 Th ete xt ur e pr of il e sp ec tr um Ac co rd in g to exp er ie nc e ga ined in our pr oj ec ts , th e fo ll owing po in ts sh oul d be ob se rve dwh en me as ur in g ro ad te xt ur e. Th eba si c me as ur ed da ta sh oul d be a lo ng it ud inal pr of il e cur ve ca pa bl e of reso lvi ng th e ma cr ot ext ur e of the sur fa ce . Th e re qui re d re solut io n is <1 m m ho ri zo nt ally an d <2 0 ;t m ve rtic al ly. Th e mi ni mum le ng th of on e pr of il e samp le sh oul d be 50 0 mm. Fo r ea chsur fa ce , th e to ta l sa mp le le ng thmus t no t be le ss th an 2 0 m,pr ef er ab ly muc h lo ng er . Th e pr of il e cur ve sh oul d be sp ec tr al an al yze dei th er by an F F T pr og ra m, a di gi ta l fi lt er or by sp ec ia l an alog fi lt er s, to gi ve th e sp ec tr a] po wer le ve l in th ir d-oc ta ve or oc ta ve ba nd s havi ng ce nt er wa ve le ng th s num er ical ly co rres po nd in g to th os e us ed in acous ti cs an d vi br at io n si gn alpr oc es si ng . RORD HR CR UT EX TU RE SP EC TR H 86 -0 5-13 SP RT IR L FR EO UE NC Y 50 0 10 00 80 2 c/ n ä lf 20 50 13 0 ?0 0 I 10 00 I T I I I I I T j T I T I »I ?-F pa ;-r' r. Ill

[2

\ u-[cmwd] \ \ '\ 01 [64 ap] 50 V. ... ... "l f, :n s 45 . BGHLIWdUH BöanBL 13031 BEHLXBL ,A SP HA åT CO NC RE TE M A l e . 4" CH IP PI NG S

35 \. AS PH AL T CO NC RE TE MA X 8 MM CH IP PI NG S 30 , 2 5 1 l I J L 1 1 I l l l l l L 1 1 1 t t i l l 1

'\.

' h ' l l 1 1 7 a 5 0 0 2 0 0 10 0 50 20 10 S 2 1 . -T E X T U R E U R U E L E N G T H ng - 5-Typ ic al th ir d-oc ta ve te xt ure sp ec tr a fo r some co mm on pave me nt typ es . Th ein ff ue nc e of th ero ad sur fa ce and it s Ch ar ac te rimt io n ll The co ve re dte xt ur e wa vele ng th ba nds sh oul d beat le as t 2-20 0 m m cent er 'a ve le ng th s -To ha vepr ac ti ca l va lues al l sp ectr al po we rle ve ls (o r 't extur e le ve ls 'as th e te rm sh oul d be in th is ap pl icat io n) sh oul d be exp ress ed in dB re la ti ve to

10

"

12

m2

:

Te xt ure le ve l =LÃ = lO lo gloñJ -E EMB) wh ere TÅ is th eme an sq uare (p owe r°) of th e prof il e cur ve (m 2) in the co ns ider ed wa ve leng th (i. )b and. TÅ co ul dal so be expre ss ed as the sq ua re of th e R M S di sp la ceme nt of the pr of il e and th e refe re nc e 10 '12 m 2 is equi va le nt to th e sq uare of th ere fe re nc e RM S di sp lace me nt of 10 '6 m. Th e mo st rele va nt te xtur e wa ve leng th ba nd sar e th os eco in ci di ngwi th th e cr it ical wa ve le ngth s as di scus se d ea rl ier. Th er ef or e, in mo st situa ti on s, itis ad equa te to le tth e te xt ur ele ve ls in two cr it ic alwa ve le ng th oc ta ve ba nds re pres en t th e nois e-re le vant te xt ur eCh ar ac te ri stic s, na me lyL8 0 (2. = 80mm ) and L5 (2. = 5mm ). Th esi tua ti on s wher e th is may be in sufii ci en t ar e when th er e is an arti ñc ia ll y prod uc ed pe ak in th e te xtur e sp ec tr umsuc h as for gr oo ve dce me nt co ncre te sur fa ces. So me typic al te xt ure pr of il e spec tr a ar e sh own in Fi g. 5. Age ne ra l feat ur e is th atth er e us el to be a maxi mum at a on e-th ir d oc tave lo ng er wave le ng th th an th e maxi mum chip pi ng si zein th e pa ve me nt . A wo rk in g group wi th in th e Gr oup of Ra pp or te ur son Br ak es and Run ni ng Ge ar(a sub gr oup wi th in theU ni te d Na tion s Ec on omic Co mm issi on fo rEur op e) ha s prop os ed a te xtur e sp ec iñc atio n us in g th ewa ve le ng th band s of 80 an d5 m m to ch arac te ri ze two re fe re nc e road sur fa ce s-on e sm oo th an d on e r o ugh -fo r tyr e/ro ad no is e meas ur em en t. The pr op os ed s ta nd ar d is il lus tr at edin Fi g. 6. 4. 3 Te xt ure me as ur in gin st rum en ts Se ve ra l ve rsio ns of co nt actl es s pr of il omet er s-ge ne ra lly ut il izi ng ala se r be am -are no w inus e. Th e two typ es at th eSwe di sh Road an d Tr af fic Re se ar ch In st it ut e are sh own in Figs 7 an d 8. The st at io na ry typ e wa s pr es en te d for th e fi rs t ti me at th e In te rnat io na l Tyr e Nois e Co nf er ence in St oc kh ol m in19 79 . Fo r thos e no t in po sses si on of an ypr oñl om et er ther e is th e poss ib il it y of prod uc in g a plas ti c mo ul di ng of th e ro ad sur fa ce an d se nd in g this fo r an al ysis to a _l abor at or y ca pabl e of pr of ilom et er me as ur em ent. ' The me th od most c o m m o n ly us ed to day fo r te xt ur eme as ur em ent is, ho we ver, th e sa nd -pat ch me thod . In th isme th od a known vo lum e of a sp ec ia lsa nd is po ured on th esur fa ce . Wi th th e he lp of a di sc th esa nd is .. .-..-.-....-...._.

12 Ul f Sa nd be rg O O [801081 u I 0 in |8^8| amixaiomeui 851303 I 0 't

3 0 .1 30 40 5 0 60 "F in e ma crot ext ur e leve l" L5 [d B] Fi g. 6. Ma cr ot ext ur e va lue s, exp re ss ed as te xt ur e le ve ls in two wa ve le ng th ba nd st asl sug ge st ed fo r two re fe re nc e ro ad sur fa ce s fo r tyr e/ ro ad no rs e me as ur em en ts . +, No mi na va lue s 83 , to le ra nc es ar oun d no mi na l va lue s. Ot he r sym bo ls re pr es en t ac tua l te st sur fa ce s. u _ 1 0 ' a n , _ .I i . _Å v 7 1 5 0 09 -_ h. , . -' ' > . . _ . . -' . u -. . P . . . A A m i ut ' w-b M " _' t . .. . J ' _ _ y, 3 . . _ I . : 'av *Mak e-;< 5 4 I. 4. :. u\ 1 -Fi g. 7. Th e st at io na ry la se rpr oñl om et er at th e Swe di sh Ro ad an dTr añi c Re se ar ch In st it ut e. Mc as ur in g sp ee d = swe cp sp ee d = 01 m/ s. T/I t' in flue nc e (i f/ liv rum ! .s ur/int ) an din' i' It t/ rarl r' ri nui zm 13 Fi g. 8. Th e mo bile la se rpr of il omet er at the Swe di sh Ro ad an d Tr af fic Re se arch In st itut e. Meas ur in gsp ee d = lOmf s (3 6km ,h ). Spre ad out in to a Ci rcul ar sp ot un ti l th esur fa ce depr es si ön sar e fi ll edto th e leve l of th epe ak s. The vo lum eof th e sand is di vi ded by th ear ea co vere d *b y th esa nd to ge t th e ave ra ge *text ur e de pth. Un fo rtun at el y. de ta il s in th e me th od as to sand gr ai nfr ac ti on ,sa nd vo lum e an ddi sc di am eter ar e not un ive rs al an d m ay di ñe r from ca se to ca se . . As me ntio ne d earl ie r. th is me th od is no t suñi ci en t to ch ar ac teri ze th e sur fa ce co mple te ly: ne ve rt he less suc ha me as urem en t (well pe rf orme d) is bett er th anno me asur em en tat all . \ Ther e is asi gn iñc ant co rr el atio n be twe en th esa nd -p atch -m ea sur ed te xt ur ede pt h an dth e sp ec tral le ve ls ofth e te xtur e. At th ete xt ur e wave le ng th of 80 m m th is re la tion fo r Swe di sh meas ur em ents wa s foun d to*h ave 'a co rr el atio n co eñi cien t of 088 an d at th e S -mm te xt ur ewa ve le ngth th e co rr el atio n co eñi cien t wa s 0-84 (4 5 data pa ir s). So , it ispo ss ib le toes ti ma te L8 0 an dLs fr om sa nd -p atch me asur em en ts. 4.4 So un dab so rpti on me asur em ent So un dab so rp tion m a y be me as ured in th eno rm al wa y by eith er th etub e me th od (K un dt 's tub e) or by th etwo -m icro ph on eme th od ba se d on F F T an alys is . Howe ve r. asth e so und ab so rpti on is negl ig ib leon sur face s ot he r *gm*- , .. ...Ui .

14 ( 'fl .b. .ll lt/ l'l t'l 't' Fi g. 9. Me as ur em en t of so un d pr op ag at io n fr om a lo ud sp ca ke r on th e ro ad sur fa ce to a mi cr op ho ne on t0 p of th e tr ip od . Th e tr ip od is co vc re d wi th ab so rb in g ma te ri al to avo rd re ñc ct io ns . th an pe rvi ous , po ro us or dr ai na ge typ es (wh at eve r th ey ar ec al le d) th is me as ur em en t is un ne ce ss ar y in mo st ca se s. A A pr ob le m is th e in fl ue nc e on th e re sul t of th e an gl e of in ci de nc ce of th e so un d. . In th e tub e me th od . fo r in st an ce . th e an gl e is 90 °, wh ic h 1.3 cl ea rl y un typ ic al of th e re al si tua ti on , On e wa y to ge t ar oun d th is pr ob le m is to us e th e two -m ic ro ph on e me th od , or to me as ur e th e di ff er en ce in so un d "p ro pa ga ti on ove r a dr ai na ge sur fa ce in re la ti on to a de ns e sur fa ce , us mg a .s ta nd ar d so un d so ur ce in pl ac e of th e te st ve hi cl e (o r tyr e) an d a me as ur mg mi cr op ho ne at th e no rm al me as ur in g di st an ce . Suc h a me as ur em en t ts il lus tr at ed in Fi g. 9. _ In th e fut ur e, a muc h si mp le r wa y to ch ar ac te ri ze th e so un d ab so rp ti on eñe ct on tr af fi c no is e mi gh t be to us e th e te nt at ive re la ti on fo un d be twe en th e . . . . l' ' no is e re duc ti on an d th e ai r vo lum e in th e dr am ag e la ye r. Th e in fl ue nc e of th e ro ad sur fa ce and its Ch ar ac te ri za ti on 15 4. 5Me ch an ic al st iñn es s orim pe da nc e me as ur em en t Fo r me asur em en t of th e pa ve me nt st if fn es s, a me th od ba se d on th e fo rc e sp ec tra of ha mm er im pa ct s onth e pa ve me nt ma y be us ed. A sm al l ha mm er , ofma ss ap pr oxi ma te ly 02 kg , with a fo rc e tr an sd uc er , is kn ocke d ag ai ns t th e pave me nt . Th e co nt ac t ar ea with th e pa ve me nt (h am me r tip) sh oul d ha ve a di am et er of 20 mm . Seve ra l im pa ct s wi th pe ak fo rc es ar oun d 30 0N sh oul d be sp ec tr al an al yze din 1/ 3-oc ta ve ba nd s. Th e en er gyle ve l of th e im pa ct sp ec tr a fo r th e freq ue nc y ra ng es of 10 0-50 0 and 50 0-16 00 H z (s um of 1/ 3-oc ta ve ba nd s) sh oul d be co mp ut ed . Th e di ff er en ce be twe en th e hi gh -fr eq ue nc y en ergy an d th e lo w-fr eq uenc y en er gy is co ns id ered to be a me as ur e of th est iñn es s. Th e pr evious ly de sc ri be d me th od ha s ce rt ai n pr ac ti ca lpr ob le ms . Th e st if fn es s in fl ue nc eon tyr e/ ro ad no is e is re la tive ly sm al l an d an al te rn at ive si mp le cl assi ñc at io n me th od ma y bead eq ua te to exp la in at leas ts om e of th e va ri atio n th at se em s to be re la te d tost iñn es s. Th is pr ovi si on al clas si ñc at io n sh oul d be : 1. Th e typ e of pa ve me nt (a Sp ha lt co nc re te or ce me nt co nc re te , et c. ). 2. The ag e of th e pa ve me nt when no is e wa s me as ur ed. If poss ib le , th ea cc um ul at ed num be r of ve hi cl e axl es pa ss in g the pa ve me nt in it s li fe ti me up to th e ti me of no is e me asur em en t sh oul d al so be st at ed. 5 R E L A T I O N B E TW E E N Tf iA FF IC NO IS EA N D R O A D C H A R AC T E R I S T I C S 'I n Se ct io n 3 th e re la ti onbe twe en tyr e/ ro ad no is e an dte xt ur e wa s de sc ri be d. But wh at is th e re la ti on be twe en tr af ñc no is e an d te xt ure? If traf ñc no is e wa s pur el y tyr e/ro ad no is e th e an swe r wo ul d cl ea r, ho we ve rco mp li ca te d. But ma tt er s are eve n mo re co mp li ca te d than th is , be ca us e in th isca se po we r-tr ai n no is e mus tal so be ta ke n in to ac co un tan d th is is no t co mp le te ly in de pe nd ent of ro ad sur fa ce Ch ar ac te rist ic s. It is "i nñue nc ed by the so un d ab so rp ti on of the ro ad an d is di ff er en t for tr uc ks and ca rs . Th e latt er pr ob le m wi ll be le ft mo ment ar il y. An exp er im en ta l prog ra mm e sti ll in prog re ss , in cl ud in g tr af hcno is e L Am me as ur em en t, ve hi cl e pa ss -b y peak no is e sp ec tr a, ve hi cle Cl as si fi ca ti on , sp ee dme as ur em en t an d ro ad te xt ur e me asur em en t, ha s co nf ir med th at th e br oa d line s in no is e-te xt ur e re lati on s pr es en te d ea rl ier fo r tyr e/ ro ad no is eho ld al so fo r no is e from fr ee -ño wi ng tr af fi c. Figur es 10 an d 11 ,wh ic hpr es en t th e re sul ts fo rf our of th e 26 te st ed ro ad s, ma y illus tr at e th is . It ca n'be se en th at a ro ughe r te xt ur e me an s ...4-4 . . i. .N . M,....w.__.___,_

l6 U1 / Sa nd be rg dB NO Rn RL IZ ED TR RF Fi C ND IS E S' PE CT RR ?E EK HT ?R SS -B Y 10 0 r vywvwwuwvvv' * vvvvi vvvvr vñvvuvwvw LI GH T UE Hl CL ES RO UG H no en (n o 43 75 0-11 9) 4 -LI GH T UE HI CL ES sn oo m RO RD (n o 30 75 0-09 9) 30 --HE RU Y UE Hl CL ES RO UG H RO HD (H 0 43 75 0-12 0) 4 " n-- - -- a-HE RU Y UE HI CL ES sn oo m Ro nn (n o 30 m -m i 80 -. 70 _. 50 .L so .. l . A L A J I . A L A A A A A A A A A 4 4 0 v* ' ^ ^ ' * ^ ^ ' v ' . I . r . 4 I . 1 I . . I . . I . . 16 31 .: 6 3 1 2 5 '2 50 5 0 0 li t ? K 4! ! av I S K 3 2 k FR EO UE NC Y [Hz] Fi g. 10 . Fr eq ue nc y sp ec tr a (p ea k pa ss -b y) fo r tr af fi c no is e of li gh t an d he avy ve hi cl es tr ave ll in g at 80 -9 0 an d 70 -8 0 km /h , re sp ec ti ve ly, on ve ry sm oo th an d ve ry ro ug h te xt ur ed sur fa ce s. dB NO Rn ñL lZ ED TR RF FI C HO IS E S' PE CT RH ?E EK RT ?H SS -B Y ' 10 0 vuvvv r vvvv1 1 w' Y T v vT va T vY Y va T V . _ _ _ -Ll GH 1 UE Hl CL ES sn oo m ns m. m 0 46 -0 75 1-11 3) l 1 -. .. .. .-. LI GH T UE Hl CL ES na nm . ns rn . m0 44 .-0 vs x-m) so .. .. .. .-. HE RU Y UE HI CL ES sn oo m ns vn . mo 46 -0 75 1-11 4) 4 "' ... ... ... ... ... . HE RU Y UE HI CL ES na nm . ns m. m 0 44 -0 75 1-10 5) 70 .... 60...L : 5 : 0 : ' li k' ' ; K v 1 : K v ' Bi k' 'l IE K' j S Z K -FR EO UE NC Y §42 ] Pi g. ll ." Fr eq ue nc y Sp ec tr a (p ea k pa ss -b y) fo r tr af fi c no is e of li gh t an d he avy ve hi cl es tr ave ll in g at 80 -9 0 an d 70 -8 0 kn i/ h, re sp ec ti ve ly, on an ol d. sm oo th . de ns e as ph al ti c co nc re te an d a ne w dr ai na ge aS ph al ti c co nc re te sur fa ce . . v ' i s 53 12 5 25 0 Th e in flue nc e of th e ro adsur fa t-e an d it s t'li ar at 'm ri sa li nn 17

[men] suv:) XBLUV-I 8

0 -78 ---1 7 5 -e Ac ce le ra ti ng tr af fic (l awvs pe ed ) .. Q L o w Sp ee d D "D ra in ag e" as ph ai t I 00 .. we t 0r wi nt er ON ew. de ns e bi tum in0us swf ac es OO ld , de ns e. aS ph al t co nc re te ' 0 00 .. we t or wi n t e r A Si ng le Sur face dr es sm g (Yi) v Do ub le Sur fa ce d r e s s i n g(7 2) 7 4 -72 I I I I I I 30 82 84 se se 90 92 ld Bt Al l LA ma xTR UC KS Pi g. 12 . Re la ti on betwe en tr af fi c no is e em is si onle ve ls (m axi mum at pa ss-b y) fo r ca rs an d he avy tr uc ks (_ 24 axl es ). Al l leve ls ar e ave ra ge s fo r al l ve hicl es in th e cl as s an d no rm alize d to 90 an d 70 km/h . re sp ec ti ve ly. Th e sq uare an d tr ia ng le sym bo ls outsi de th e do tt ed ar ea s ar e ro ad sur fa ce s wh ic hha ve ch an ge d a lo t fr om thei ro ri gi na l st at ea nd no war e in th e la te rp ar t of th ei r surfa ce lif e. an incr ea se d no is e ga t low fr eq ue nc ie s but decr ea se d no is e at hi gh fr eq uenc ie s. Fo r tr uc ks , th efo rm er st at em en t ha ssm al le r sup po rt owi ng to an inñue nc e fr om po we r-trai n no is e in th is ra nge al so . lt ca n be ob serve d th atth ed at a ar e ave ra ge spec tr a fo ra no rm al ize dmi x of ve hi cl es ad juste dt o Swe di sh tr af fi c co mp os itio n an d ba se d on me as urem en ts on a va ri et y of in di vi dua l pa ssin g ve hi cl es .6 Figur e 11 il lus tr at es the ef fe ct of us in g a drai na ge sur fa ce in st ead of a de ns e, Sm ooth sur fa ce . Th eef fe ct s se en ma y be a co mb in at io nof th e re duc ti on ofth e hi gh -f re que nc y tyre/ ro ad no is e-ge ne ra tion me ch an is m an d i th e di ff er en ce inso un d ab so rp ti on . Inth e fo rm er ca se , th eno is e-re duc ti on ef fe ct may be pr es en t at lo we rfr eq ue nc ie s fo rt ruc ksth an fo rc ar s, be ca us eof th e di ff er en tcr os so ve r fr eq uenc ie s. As di sc us se d ea rl ier, a ce rt ai ns ur fa ce may be re la ti ve ly no is yfo r ca rs but si le nt fo r tr uc ks . Fi gure 12 sh ows th at th isis tr ue in ce rt ai n ca se s.Ho We ve r, dr ai nage sur fa ce s ap pe arto be si le nt fo r bo thve hi cl e ca te go ri es. Fi gur e 12 al sore ve al s th at th e re pe atab il it y of th e me as urem en ts is qui te go od asth e Sp re ad in nois e le ve ls wi th in ace rt ai n ve hi cl e/ road typ e is sm al l, wi th the exc ep ti on of drai na ge pa ve me nt s, fo r wh ic h it iskn own th at th e ._ i...

00/ UI (WW-*l (HI '06) () .wlan/n.: Annat] 0,, 00/ 0/ ('I/tw/ oz ) .va/.2.111.141 .than 0" 00]

(('y mp ) " V'y '[000] unpzvgm.) omm u; anuvuqyia UI

(till/ DI Og)

snyggt/.m .t'aøm/ %

poods put: *sepiqu

ÅacaH JO uoguod aqi uo oauapuadoa sig put: 'SODBJJHS peou mamma uaamtag ' V'I *laAaq Kq-ssed uogssgtug asgoN ut aauaiaggg panseaw Z H'IHVJ.

(131m popnpug aq

gm mnp :now se 'DAlmlUOI om sonpm "V 'paods tumsuoa m omm] ilugmouoaid 'OOUOJOJDJ om om (| '0N) sooejms OIDJDUOD tlnque osuap 'qtoowy sam' pvo.:

_[0 JaqtunN dmub' aimfms pøoy

Ul f Sa nd be rg Th e in fl ue nc e of th e ro ad sur fa ce an a' il : Ch ar ac te ri za ti on 19 ° __ exa ct de si gn an d co nd it io n isve ry cr it ic al .l ta ls o gi ve s th e hi nt -a s a re li efto th e ho pe le ss ne ss co nc er ni ng th e co mpli ca te d re la ti on s be twe en no is e an d ro ad sur fa ce -t ha t it ma y be po ss ib le af te ra ll togr oup to ge th er ce rt ai n ma in 't yp es of ro ad s wi th re sp ec t to ove rall A-we ig ht ed no is e le ve l. Ho we ve r, th e li mi ta ti on sa re sti ll th at th es er oa dswi th in th e typ e mus t be qui te si mila r an d th at a di ff er en ti at ion is ma de wi th re sp ec t to ve hi cl etyp e an d sp ee d. A pr ovi si on al ta bl e ba se don th is is pr es en te d in Ta bl e 2. It sh oul d be co ns id er ed as an exa mp le. Th em ai n co nc lus io n is th at th e tr af li c no is e co rr ec ti on owi ng to th e road sur fa ce is a va ri ab le of ve hi cl e sp ee d, _p 0r ti on of he avy ve hi cl esan d-of co ur se -t he ro ad sur fa ce typ e an d co nd it io n. Ifth e re fe re nc e ro ad sur fa ce is a de ns e, sm oo th -t ext ur ed as ph al t co nc re te th e co rr ec ti on sma y be up to +6 dB (A ) (i n mo de ra te we t co nd it io ns at lo w sp ee ds ) an d do wn to -5 dB (A ) (f or dr y, dr ai na ge as ph al t) . Th en ,0 nl y sur fa ce s c o m m o nin Swe de n ar e in cl ud ed ; exc lud ed ar e sur fa ce s suc h as ce me nt co nc re te and pa vi ng st on es wh ic h, as ar e kn own fr om ot he r st ud ie s, ca n be ve ryno is y. . Di sc us si on ab out ro ad sur fa ce in fl ue nc e ha s be en li mi te d to fr ee -ño wi ng tr afñc at 50 km /h an d at hi gh er sp ee ds . Fo r in te rr up te d-ño w traf ñc th e tyr e/ ro ad no is e in ñue nc e is, so fa r, m uc h lo we r an d th us th ero ad sur fa ce co rr ec ti on mus t be lo we r, if an y. Th en it shoul d be ob se rve d th at th e so un d ' ab so rp ti on pr op er ti es ar e in ñue nc in g al sopo we r-tr ai n no is e so at le as t th is ro ad sur fa ce co rr ec ti on ma y no t be ne glec te d; ho we ve r, at pr es en t it ca nn ot be qua nt iñe d. 1+ + 2+ 3+ 1+ + {+ + |+ + 75+ 17+ '9+

t:

+ 1+ + 2+ S+ tri-9+ E+ 2+ 6F U T U R E T R E N D S Ls 'cm se *ll vs *E5 'zs 'ow'w'9s'ss LS |S 09Vzv St'om ZW Z9t' IW [917 Th e im po rt an ce of co rr ec ti ng tr af fi c nois e pr ed ic ti on s fo r ro ad sur fa ce in ñue nc e is li ke ly to in cr ea se in th e futur e as a re sul to ft he fo ll owi ng ma jo r re as on s: ^ ' 1. Th e pr ed ic ti on mo de ls ar e re ñn ed 't o suc h an ac cur acy th at sm al le r sys te mati c de vi at io ns wi ll ha ve an in cr ea se d re la tive in fl ue nc e. 2. Di ff er en ce s in ro ad an d tyre co nd it io ns (d ry/ we t, wi nt er ) wi ll be co me mo re an d mo re in te re st in g in li ne wi th th ein cr ea se d de ma nd s fo r ac curac y. 3. Owi ng to th e ti ghte ne d ve hi cl e no is e em is si on li mi ts , as agre ed in Eur Op ef or th e la tt er pa rt of th is de ca de , th e re la ti ve im po rt an ce of tyr e/ ro ad no is e wi ll in cr ea se si nc e th e ne w em is si on li mi ts ar e in ñuenc in gwo nl y po we r-tr ai n no is e. 4. Owi ng to no is e-re ducti on re qui re me nt s, dr aina ge sur fa ce s wi ll be us ed mo re co mm only wh ic h wi ll re quire a re la ti ve ly la rg eno is e co rr ec ti on . 'mmm-z] *[Å ;dm '30!8891p ;otaan aiâugs 'g

qânm tuntpatu 'mur g-p 'Ja/(Bl

'Mau apmuoa tlequu asuap *qtoows 'z .L9IHVH *utavw '33

*UGQVH ad/(i *uequB aäeugeig '{

.L9IHVH :Latavw '39 q8n01 utngpam JO qânOJ *utsam pappnts)

:mm 'I 'ON om 'L suopgpuoo JalulM ut tnq { 'ON mm 'Ol

:salsa- (1 Poppms)

max [-0 'a '191 ut pagpads s; *paseq s; alqm sgqt qogqm uo 'dnorâ top!qu Menu, aut ugqtgm sath :43ng tuaiaygp JO uotuod aql,

puooas 'ZA ;ad/(1 'Suissaip aoeyns álqnog 'y

'spaads Jaqâiq nzqmawos in stuatuomseatu mot; p918|0dm1xa :m2 Kau) se ugeuaaun aJe q/uni 05 18 san|eA gul"'mous ,(q âuguomas pua uoudJosqe a|qgâg|80u uo poseq am sanleA :nu 'suompuoa .mugm u; u,

suougpuoa JOIUlM ut inq *l 'ON :mn '6

'129,4 [2 'amauoo 1|equa asuap utoows '1

1% m 'i'ON min '8

20 Ul f Sa nd be rg 5. At te mp ts to ma ke th e tr af fi c ño w sm oi ot he r, th e bui ld in g of mo re hi gh -s pe ed ro ad s an d th e im pr ove d ve hi cl e de si gn wi ll pr ob ab ly me an hi gh er ave ra ge sp ee ds wh ic h wi ll in cr ea se th e re la ti ve im po rt an ce of tyr e/ ro ad no is e. 7 C O N C L U S I O N S Fo r fr ee -H owi ng tr af fi c it ha s be en sh own exp er im en ta ll y th at th e ro ad sur fa ce in ñue nc e on th e A-we ig ht ed tr af fi cn oi se le ve l is at le as tup to + 6 dB an d do wn to -5 dB re la ti ve to a dr y, _s mo ot h, de ns e as ph al t co nc re te pa ve me nt . Th is wa s me as ur ed as typ ic al fo r Swe di sh ro ad s; al so , fr om kn own st at is ti cs of ro ad s in th eo th er No rd ic co un tr ie s it ma y be ge ne ra li ze d to th es e co un tr ie s as we ll . Ext re me sur fa ce s suc h as pa vi ng St on es an d gr oo ve d ce me nt co nc re te ar e li ke ly to in cr ea se th is ra ng e eve n mo re . Th e ab ove ca ll s fo r th e in cl us io n of a co rr ec ti on te rm fo r ro ad sur fa ce in th e pr ed ic ti on mo de ls fo r tr af fi c no is e. Exp ec te d tr en ds ove r th e co mi ng de ca de s wi ll eve n mo re sup po rt th is vi ew. Ho we ve r, it ha s be en fo un d th at th e co rr ec ti on te rm is a va ri ab le de pe nd in g on ve hi cl e typ e (l ig ht or he avy) , ve hi cl e sp ee d an d-of co ur se --ro ad sur fa ce typ e. It ma ke s th e co rr ec ti on qui te co mp li ca te d wh en hi gh ac cur ac y is de si re d. To ma ke th in gs eve n mo re co mp li ca te d, th e co rr ec ti on te rm fo r A-we ig ht ed tr añi c no is e le ve l ma y be qui te di ff er en t fr om th at re qui re d to co rr ec t th e no is e le ve ls wi th in ce rt ai n fr eq ue nc y ba nd s. Fo r in st an ce , no is e di st ur ba nc e in do or s wi ll be mo re af ie ct ed by th e lo we r fr eq ue nc ie s th an th e A-we ig ht ed le ve l me as ur ed out do or s. No is e co rr ec ti on owi ng to ro ad sur fa ce s is ve ry di ñe re nt at lo w fr eq ue nc ie s fr om th at ne ce ss ar y at hi gh fr eq ue nc ie s. _ ' Fo r do cum en ta ti on of ro ad sur fa ce Ch ar ac te ri st ic s in tr af fi c no is e me as ur em en ts , it is re co mm en de d to me as ur e th e fr eq ue nc y sp ec tr um of th e lo ng it ud in al te xt ur e pr oñl e cur ve . Th ir d-oc ta ve or oc ta ve -b an d re so lut io n is fo un d to be ad eq ua te . F or 't hi s pur po se th er e ar e co nt ac tl es s pr of il om et er s, but wh er e suc h in st rum en ts ar e no t ava il ab le th ea lt er na ti ve ,s im pl e wa y is to ma ke a *p la st ic ' mo ul di ng of th e sur fa ce te xt ur e wh ic h ma y be se nt fo r an al ys is to an ot he r la bo ra to ry. __

lf

th

e

rO

ad

sur

fa

ce

is

of

th

ed

ra

in

ag

e

typ

e.,

'a

dd

it

io

na

lC

ha

ra

ct

er

iza

ti

on

of.

its ac ous ti c im pe da nc e is ne ed ed . A me as ur em en t of th e so un d ab so rp ti on co ef ñc ie nt or so un d pr op ag at io n re la ti ve to an ac ous ti ca ll y h ar d sur fa ce as a fun ct io n of fr eq ue nc y is re co mm en de d. Id ea ll y, a sup pl em en ta ry me as ur em en t of th e me ch an ic al im pe da nc e at ce rt ai n fr eq ue nc ie s sh oul d be ti ng ad e to ch ar ac te ri ze th e st if fn es s. Ho we ve r, at pr es en t it ma y be suf fic ien ti" jus t to cl as si fy th e sur fa ce typ e, its ag e an d its exp os ur e to tr añi c. ' Th e in fi ue na ' of th e ro ad sur fa t' an á' it sC ha ra ct er iza ti on Zl Th e co mp li ca te d pa tt er n de sc ri be tf ot r ma d 's ur fa ce .c or re ct io n is a co ns eq ue nc e of th e co mp li ca te d tyr e/ *r dâi ge ne ra ti on me ch an is ms as we ll as th e in fl ue nc e on so un d pr op ag at io n it ih es oun d-.a bs or pt io n ob ta in ed in dr ai na ge sur fa ce s. Fo r tyr e/ ro ad no is ti här e ar e at le as t two in de pe nd en t no is e-ge ne ra ti on me ch an is ms wh ic ha nye co ni ii ct in g re la ti on s Wi th ro ad sur fa ce te xt ur e. Th en th e ove ra ll nO iS ii E al sup er po si ti on of th e no is e fr om th es e two me ch an is ms wh ic h th en -is al lt se da n of ro ad , Sp ee d an d tyr c (a nd th us al so ve hi cl e) . Eve n if th es e me cl ayg ms an d th ei r re la ti on wi th ro ad te xt ur e ar e st ra ig ht fo rwa rd an d lo gi ca l, tl en et eH ec t fo r th os e in te re st ed in ob ta in in g si mp le re la ti on s be twe en tra ffi c no is e an d ro ad de si gn is so me th ing re se mb li ng ch aos. Ho we ve r, wh en li mi ti ng th e am bi ti øg , :p tr af fi c no is e le ve ls an d to th e ro ad sur fa ce s mo st typ ic al oå th e 7* "C co un tr ie s it ap pe ar s jus ti ña bl e to pr od uc e a ta bl e of co rr ec ti gm fo r di lT er en t ro ad sur fa ce typ es pr ovi de d it is di ff er en ti at ed fo r ve hi cl e mi n an ds pe ed . Suc h a ta bl e is, in fa ct , th e on ly pr ac ti ca l so lut io n to th e pr ob le m: of pr ed ic ti ng th e tr af fi c no is e le ve l be fore th e ro ad sur fa ceis co ns tr uc te d. i Th en , af te r al l, out of th e ch ao s ma y :c em er so me or de r. RE FE RE N'CE S l. Ca lc ul at io n of Ro ad Tr af ic No is e, De pa rt pa nt of th e Envi ro nm en t, WC lS h Of ñce, 19 75 . -2. Bere kn in g va n We gve rkee rs ge lui a' , Mi ni ng -; ri eva n Vo lk sg czo nd he id en Mi lj euh yg ie ne, Th e Ne th er la nd s, 1981 . . ' 3. Ac ous ti cs -M et ho ds fo r the Me as ur em en t of ' Road Tr af ic Na is ei Aus ua li an St an da rd 27 02 -1 98 4, St an dard s As So ci at io n of Aus tr al ia . 1984 -4. Th e Co mp ut in g M od el fo r Ro ad Tr af ic No is e. , St at en s/ pl an ve rk , Re po rt No . 48 , St oc kh ol m, 19 80. i 5. U. Sa nd be rg , Wi ll ti re /r oa d no is e li mi t fut ur e 've hi cl e n0 is e re duc ti on s? , Pr oc . Inte r-No is e8 2, Sa n Fran ci sc o, pp . 12 7-30 . ' 6. U. Sa nd be rgan d J. We näl l, Th e In fl ue nc eor. Tr af ic ' NO ÃY e Of Ro ad s an d In di vi dua l Ve hi cl e No is e Em is si on s, Swe di sh Ro ad an d Tr af fi c Re se ar ch In stitut e, Li nk oepi ng , 19 85 . i 7.U. Sa nd be rg an d G. De sc or ne t, Ro ad sur fa ce in li ue nc e on ti re /r oa d no is e. Pa rt I. G. De sc or ne t an d U. Sa nd be rg , Rc ad sur fa ce in ñue nc e on ti re /r oa d no rs e. Pa rt II. Pr oc . In te r-No is e 80 , Mi am i, Ap p. 26 1-72 . 8. U. Sa nd be rg an d G. De sc or ne t, De sc ri pt io n of Me tl l' od ol og y Co nc er ni ng th e Id en ti fi ca ti on of Two Ge ne ra ti on Me ch an is ms fo r Ti re /R oa a' No is e an d th e Di fe re nt ia ti on of th e No is e in to th e Two Me ch an is ms , VT I me dd el an de No . 32 5. Swe di sh Ro ad an d Traf ii c Re se ar ch In st it ut e, 19 81 9. T. G. Cl ap p an d ..A . C. Eb er ha rd t, Sp ec tr al co rr el at io n te ch ni que s ap pl ie d to eva lua te no is e an d sa fe ty tr ad eo ff s in ti re /p ave me nt inte ra ct io n, Tr an sa ct io ns of th e A S M E : Jo ur na l of Vi brat io n, Ac ous ti cs . St re ss and Re li ab il iU " i" Da ig ni No . 83 -D ET -l O, Am er ic an So c. of Me ch . En g. , Ne w Yo rk . NY .19 83 -.-.. 0... .--.---. ..

22 10. 11. 12. 13. Ul f Sa nd be rg K. Sc ha af an d D. Ro nn eb er ge r, No is e ra di at io n "f ro m ro Il in g "t ir es -s oun d am pl iñc at io n by th e ho rn -e ff ec t, Pr oc . In te r-No is e 82 ,S an Fr an ci sc o, pp . 13 1-4, U. Sa nd be rg , Re duc ti on of ti re /r oa d no is e by dr ai na ge as ph al t, Re if en ge råuA 'c -h un d St ra ss en ba u, In te rn . Se mi na r, Mi tt ei lun g N r 57 , In st it ut für St ra ss en -. Ei se 'n ba hn -un d Fe ls ba u, ET H, Zür ic h, 19 84 , pp . 18 7-20 5. U. Sa nd be rg , A -H _t 'p or he si s C on ce rm 'n g th e In fl ue nc e of Pa ve me m Sl ij ñze ss on Ext er na l Tf re /R oa d No is e, Swe di sh Ro ad an d Tr af fi cR es ea rc h In st it ut e, 19 80 . U. Sa nd be rg an d J. A. Ej sm on t, Re fe re nc e ro ad sur fa ce s fo r us e in ti re /r oa d no is e me as ur em en ts , Pr oc . In te r-No is e 85 , Mun ic h, pp . 12 19 -2 2. 14 . G. De sc or ne t an d U. Sa nd be rg , Pr op er ti es . of Dr ai na ge As p/ :alt Im po rt an t fo r Tr af ic No is e Re duc ti on , Be lg ia n, R' oa d Re se ar ch Ce nt er /S we di sh Ro ad an d Tr af fi c Re se ar ch In st it ut e, 19 86 /i n pr ep ar at io n. t... i.