May 2010
Reprint Report: UCPRC- RP- 2010- 02
Temperature Influence on
Road Traffic Noise:
Californian OBSI
Measurement Study
Authors:
Hans Bendtsen, Danish Road Institute— Road Directorate;
Qing Lu, UC Pavement Research Center; and
Erwin Kohler, Dynatest Consulting, Inc.
This report is based on research performed by the Danish Road Institute- Road Directorate
on behalf of the University of California Pavement Research Center for the California
Department of Transportation, and is reprinted here in its original form.
Work Conducted as Part of the “ Supplementary Studies for Caltrans QPR Program” Contract”
PREPARED FOR:
California Department of Transportation
( Caltrans)
Division of Research and Innovation
PREPARED BY:
The Danish Road Institute—
Road Directorate and
University of California
Pavement Research Center
Danish Road Institute
ii UCPRC- RP- 2010- 02
DOCUMENT RETRIEVAL PAGE Reprint Report
UCPRC- RP- 2010- 02
Title: Temperature Influence on Road Traffic Noise: Californian OBSI Measurement Study
Author: H. Bendtsen, Q. Lu, and E. Kohler
Prepared for:
Caltrans
FHWA No.:
CA101735B
Date Work Submitted:
July 2009
Date:
May 2010
Contract/ Subcontract Nos.:
Caltrans Contract: 65A0293
UC DRI- DK Subcontract: 08- 001779- 01
Status:
Final
Version No:
1
Abstract: The work described in this report is adjunct to a five- year study of tire/ pavement noise undertaken by the University
of California Pavement Research Center for the California Department of Transportation under the Partnered Pavement Research
Center program ( PPRC). This part of the study was performed in cooperation with the Danish Road Institute/ Road Directorate, and it
examined the influence of air temperature on tire/ pavement noise measurements performed on two types of tires ( Aquatred and
Standard Reference Test Tire [ SRTT]) on different asphalt pavement surfaces using the On- board Sound Intensity ( OBSI) method.
Field noise measurement testing was carried out in two series: one in the Southern California desert on State Route 138 using the
SRTT, and the other with data collected on a statewide selection of pavements tested with the Goodyear Aquatred tire in an earlier
part of the PPRC noise study. The field measurements yielded data for deriving air temperature coefficients for the two types of tires,
and a comparison of them is made.
A worldwide survey of the available literature accompanies the field work and analysis, and a summary of it is used to compare
the air temperature coefficients of the SRTT with a combination of tire types used in European testing. In addition, findings in the
literature serve as the basis for a series of predicted temperature coefficients for passenger cars on various cement concrete and
asphalt pavements.
Finally, the report presents ten general conclusions drawn regarding the relationship between air temperature correction and
tire/ road noise on asphalt and concrete pavements.
Keywords: Tire/ pavement noise, On- board sound intensity, Temperature influence, Temperature coefficient
Proposals for implementation: It is recommended that Caltrans begin using the temperature corrections noted in this
report in its measurements.
Related documents:
• Bendtsen, H. 2009. Highway Noise Abatement: Planning Tools and Danish Examples. Reprint report: UCPRC- RP-
2010- 03
• Bendtsen, H. 2009. Noise Barrier Design: Danish and Some European Examples. Reprint report: UCPRC- RP-
2010- 04
• H. Bendtsen, H., Q. Lu, and E. Kohler. 2009. Acoustic Aging of Asphalt Pavements: A Californian/ Danish
Comparison. Reprint report. UCPRC- RP- 2010- 01
• Q. Lu, E. Kohler, J. T. Harvey, and A. Ongel. 2009. Investigation of Noise and Durability Performance Trends for
Asphaltic Pavement Surface Types: Three- Year Results. Research report: UCPRC- RR- 2009- 01
• E. Kohler. 2010. Quieter Pavement Research: Concrete Pavement Tire Noise. Research report: UCPRC- RR-
2010- 03
Signatures:
Hans Bendtsen
1st Author
DRI- DK
John T. Harvey
Technical Review
UCPRC
John T. Harvey
Principal Investigator
UCPRC
S. David Lim
Contract Manager
Caltrans
UCPRC- RP- 2010- 02 iii
DISCLAIMER
This report is based on a subcontract research study performed by the Danish Road Institute- Road Directorate
( DRI- DK) on behalf of the University of California Pavement Research Center ( UCPRC) for the California
Department of Transportation ( Caltrans). The contents of this report reflect the views of the authors and DRI-DK
who are responsible for the facts and accuracy of the data presented herein. The contents do not necessarily
reflect the official views or policies of the UCPRC, the State of California or the Federal Highway
Administration. This report does not constitute a standard, specification, or regulation. The content of the
original is unchanged in this version and has been reprinted with the consent of DRI- DK.
For more information:
University of California Pavement Research Center, Davis
One Shields Avenue, Davis, CA 95616
University of California Pavement Research Center, Berkeley
1353 S. 46th St., Bldg. 452, Richmond, CA 94804
www. ucprc. ucdavis. edu
Danish Road Institute
Report 169
2009
Hans Bendtsen
Qing Lu
Erwin Kohler
Temperature infl uence
on road traffi c noise
Californian OBSI measurement study
xx
3
Contents
Executive summary ....................................................................................................... 5
Sammenfatning.............................................................................................................. 8
Preface ........................................................................................................................ 11
Forord ........................................................................................................................ 13
1. Introduction and existing knowledge ...................................................................... 15
1.1 The “ Tyre/ Road Noise Reference Book”.......................................................... 15
1.2 Temperature and different noise measurement methods................................... 17
1.3 Results from the German- Dutch Sperenberg project ........................................ 21
1.4 Semi- generic temperature correction method ................................................... 25
1.5 Previous American investigations..................................................................... 26
1.6 Results from a French experiment .................................................................... 30
1.7 The European Union Tire Noise Directive........................................................ 32
1.8 The challenge .................................................................................................... 33
2. The test sections ...................................................................................................... 35
2.1 The LA138 pavements ...................................................................................... 35
2.2 Californian pavements....................................................................................... 38
3. The OBSI measurement method.............................................................................. 41
4. The LA138 measurements....................................................................................... 45
4.1 Air temperature and noise ................................................................................. 46
4.2 Pavement temperature and noise....................................................................... 51
5. The California measurements.................................................................................. 55
6. Discussion and conclusion....................................................................................... 59
References ................................................................................................................... 65
4
5
Executive summary
International experience indicates that temperature is a factor which has some influ-ence
on the results of measurements of road traffic noise. The objective of this report
is to analyze how temperature affects the On- Board Sound Intensity ( OBSI) measure-ments
of tire/ pavement noise. The results are also relevant for the Close Proximity
method ( CPX) if a Standard Reference Test Tire ( SRTT) is utilized. It can be dis-cussed
whether the temperature coefficient shall be given in relation to the air, pave-ment
or tire temperature. There has so far been some international tendency to use air
temperature as an independent variable so this will be done in the following.
The work presented in this document was done by analyzing two sets of measurement
data. A series of detailed OBSI noise measurements with the SRTT were performed
on the Caltrans test sections at highway LA138 in the Mojave Desert in Southern Cali-fornia.
The measurements were carried out in the desert in wintertime where the varia-tion
of the air temperature over the day was from to 2 to 22° C. The noise has been
measured on the same day or within a few consecutive days with the same equipment,
by the same operator, and on the same pavements, at low ( morning), medium ( mid-day),
and high ( afternoon) temperatures. This ensures that the only main variable pa-rameter
during these measurements was the temperature. In the second measurement
series, a Goodyear Aquatred tire was used, which was the former standard test tire for
OBSI. The variation of the pavement temperature over the day was from to 11 to
35° C.
The objective was to perform measurements where the only variable was the tempera-ture
and where the following factors were constant:
• Same measurement tire.
• Same inflation and rubber hardness of the measurement tire.
• No changes in age, tear and wear of the measurement tire.
• Same acoustical measurement equipment.
• Measurement tire mounted on the same car.
• Same measurement operator.
• No changes in pavement conditions other than the temperature.
The coefficients of noise vs. temperature measured with the SRTT and the Aquatred
tire are significantly different. The average air temperature coefficient for the
Aquatred tire is three times higher than for the SRTT depending on the pavement type.
This means that the Aquatred tire is much more sensitive to temperature than the
SRTT. The tire hardness was a little lower for the SRTT than for the Aquatred tire ( 67
versus 69 Shore A). This might partly explain this difference in temperature coeffi-cients
but other tire properties like the chemical composition of the rubber, the tread
pattern and the tread depth differences etc. might also play a role.
6
The results from different international measurement series are also summarized in
the report. SRTT have significantly lower temperature correction factors than the other
tires and tire populations included in the comparison. This shows that the SRTT is not
very sensitive to temperature variations. The average air temperature coefficient for
the SRTT on asphalt concrete pavements is - 0.027 dB/° C. There is no big difference
between dense and open graded pavements: - 0.029 dB/° C versus - 0.026 dB/° C.
Therefore it is suggested that - 0.027 dB/° C be used as the air temperature correction
factor for the SRTT used on asphalt pavements. Third octave band correction factors
have also been determined. There has not been any data available to evaluate the tem-perature
correction coefficient for the SRTT used on cement concrete pavements.
A series of rough general average air temperature coefficients for passenger cars at
the different pavement types are predicted and shown in the table below. These coeffi-cients
are predicted on the background of the results from the different international
measurement series summarized in the report and the measurements carried out in this
project. There is no big difference between temperature corrections for dense (- 0.061
dB/° C) and open graded asphalt pavements (- 0.052 dB/° C). The correction factor for
cement concrete pavements is - 0.043 dB/° C and lower than for asphalt concrete
pavements.
Dense
asphalt pavements
( DGAC)
Open graded
asphalt pavements
( OGAC)
Average all
asphalt pavement
types
Cement concrete
pavements
- 0.061 dB/° C - 0.052 dB/° C - 0.057 dB/° C - 0.043 dB/° C
These general correction factors are relevant in relation to measurement methods
where a large amount of different light vehicles and tires are included like the Statisti-cal
Pass- By method or LAeq measurements. Generally these results are quite close to
the coefficient of - 0.05 dB/° C for passenger cars commonly used in Denmark and the
Netherlands, and to the coefficient used in the EU tire noise directive of - 0.06 dB/° C
up to 20 ° C. These factors are approximately double of those for the SRTT on asphalt
tested in California.
The following general conclusions can be drawn regarding temperature corrections to
tire/ road noise measurements:
• The air temperature has an important influence on the tire/ road noise measurements
results.
• The dependence of tire/ road noise on temperature can be approximated by a linear
relation.
• The temperature coefficient varies significantly for different tire types.
• The temperature coefficient is generally smaller for truck tires than for
passenger car tires.
7
• At low frequencies, the temperature coefficient is low. At frequencies above 1000
Hz the temperature coefficient is higher.
• The temperature coefficient is different for different pavement types.
• The temperature coefficient seems to be higher for dense asphalt concrete than for
open/ porous asphalt pavement.
• The temperature coefficient seems to be lower for cement concrete pavements than
for asphalt concrete pavements.
• The difference in temperature coefficients for different asphalt pavement types al-most
vanishes when many different tires are included.
• Temperature coefficients have to be determined specifically for each measurement
method taking into consideration the specific test tire( s) or the tire population in-cluded
in the measurements.
8
Sammenfatning
Internationale erfaringer viser, at temperaturen er en faktor som har en vis indflydelse
på resultaterne af målinger af vejtrafikstøj. Formålet med denne rapport er at analyse-re,
hvorledes temperatur influerer på On- Board Sound Intensity ( OBSI) målinger af
dæk/ vejstøj. Resultaterne er også vigtige for Close Proximity metoden ( CPX), såfremt
et Standard Reference Test Tire ( SRTT dæk) anvendes.
Arbejdet, der præsenteres i denne rapport blev udført ved at analysere to sæt måledata.
En række detaljerede OBSI støjmålinger med SRTT dæk blev udført på Caltrans’
( vejdirektoratet i Californien) prøvestrækninger på LA138 i Mojave ørkenen i det syd-lige
Californien. Målingerne blev foretaget i vintermånederne, hvor variation af luft-temperaturen
om dagen var fra 2 til 22 ° C. Støjen blev målt den samme dag eller inden
for et par efterfølgende dage med det samme udstyr, den samme operatør, og på de
samme belægninger, ved lav formiddagstemperatur, mellem middagstemperatur og
høje eftermiddagstemperaturer. Dette sikrer, at den eneste vigtigste variabel i løbet af
disse målinger er temperaturen. I den anden måleserie anvendtes et Goodyear
Aquatred dæk, som var det tidligere standard testdæk for OBSI metoden. Variationen i
belægningens temperatur i løbet af dagen var fra 11 til 35 ° C.
Formålet var at udføre målinger, hvor den eneste variabel var temperaturen, og hvor
følgende faktorer var ens:
• Samme måledæk
• Samme tryk og gummihårdhed af måledækket
• Ingen ændringer i alder og slitage af måledækket
• Samme akustisk måleudstyr
• Måledækket monteret på den samme bil
• Samme måleoperatør
• Ingen ændringer i slitagen af vejbelægningerne.
Målingerne med SRTT dæk og Aquatred dæk viser en markant forskel i temperatur-koefficienterne
for disse to dæk. Den gennemsnitlige lufttemperaturkoefficient for
Aquatred dæk var 3 gange højere end for SRTT afhængig af belægningstype. Dette
betyder, at Aquatred dækket er langt mere følsomt over for temperatur end et SRTT.
Dækkets hårdhed var lidt lavere for SRTT end for Aquatred dækket ( 67 mod 69 Shore
A). Dette kan delvis forklare forskellen i temperaturkoefficienterne, men andre dæke-genskaber,
så som den kemiske sammensætning af gummiet, slidbanemønsteret og
mønsterdybden osv. kan spille en rolle.
9
Resultaterne fra forskellige internationale måleserier er sammenfattet i rapporten.
SRTT har markant lavere temperaturkoefficienter end de andre dæk og dækgrupper,
som indgår i sammenligningen. Dette viser, at SRTT dæk ikke er særlig følsomme
over for temperaturvariationer. Den gennemsnitlige lufttemperaturkoefficient for
SRTT på asfaltbetonbelægninger var - 0,027 dB/° C. Der var ingen stor forskel mellem
tætte og åbne belægninger: - 0,029 dB/° C mod - 0,026 dB/° C. Derfor er det foreslået at
anvende - 0,027 dB/° C som den lufttemperaturkoefficient for SRTT dæk som bruges
på asfaltbelægninger. Tredjedel oktavbånd korrektionsfaktorer er også blevet fastlagt.
Der findes ingen data til at vurdere lufttemperaturkoefficienten for SRTT dæk brugt på
betonbelægninger.
På baggrund af resultaterne fra de forskellige internationale måleserier, som er sam-menfattet
i rapporten, og de målinger, der er gennemført, er en række gennemsnitlige
lufttemperaturkoefficienter for personbiler på de forskellige belægningstyper beregnet
( se tabellen nedenfor). Der er ingen stor forskel mellem lufttemperaturkoefficienten
for tætte (- 0,061 dB/° C) og åbne asfaltbelægninger (- 0,052 dB/° C). Lufttemperaturko-efficienten
for betonbelægninger er - 0,043 dB/° C og lavere end for asfaltbetonbelæg-ninger.
Tætte
asfaltbelægninger
Åbne
asfaltbelægninger
Gennemsnit alle
asfaltbelægninger
Betonbelægninger
- 0,061 dB/° C - 0,052 dB/° C - 0,057 dB/° C - 0,043 dB/° C
Disse generelle lufttemperaturkoefficienter er relevante i forhold til målemetoder, hvor
en stor mængde forskellige lette køretøjer og dæk er medtaget, som ved Statistical
Pass- By- metoden ( SPB) eller LAeq målinger. Generelt er disse resultater ganske tæt på
koefficienten - 0,05 dB/° C for personbiler, der almindeligvis anvendes i Danmark og
Holland, og koefficienten, der anvendes i EU dækstøjdirektivet af - 0,06 dB/° C op til
20 ° C.
På baggrund af dette projekt kan følgende generelle konklusioner drages angående
lufttemperaturkoefficienter ved dæk/ vejstøjmålinger:
• Temperaturen har en vigtig indflydelse på dæk/ vejstøjmåleresultater.
• Der er en lineær afhængighed mellem temperatur og dæk/ vejstøj.
• Temperaturkoefficienten varierer betydeligt for forskellige dækmodeller.
• Temperaturkoefficienten er generelt mindre for lastbildæk end for dæk til personbi-ler.
• Ved lave frekvenser er temperaturkorrektionskoefficienten lav. Ved frekvenser
over 1000 Hz er temperaturkorrektionskoefficienterne højere.
• Temperaturkoefficienten varierer efter belægningstyperne.
• Temperaturkoefficienten synes at være højere for tæt asfaltbeton end åb-ne/
drænasfaltbelægninger.
10
• Forskellen i temperaturkoefficienten for forskellige typer asfalt forsvinder næsten,
når mange forskellige dæk er inkluderet.
• Temperaturkoefficienten synes at være lavere for betonbelægninger end asfaltbe-tonbelægninger.
• Temperaturkoefficienten skal fastsættes specifikt for de enkelte målemetoder under
hensyntagen til de specifikke testdæk eller dækgrupper, som indgår i målingerne.
11
Preface
International experiences indicate that temperature is a factor which has some
influence on the results of measurements of road traffic noise. The On Board Sound
Intensity ( OBSI) method is used by University of California Pavement Research Cen-ter
( UCPRC) as well as by other researchers and consultants in USA to perform de-tailed
measurements of tire noise emission from road pavements. The OBSI method
is frequently used in noise projects performed for the California Department of Trans-portation
( Caltrans). An Expert Task Group organized by the U. S. Federal Highway
Administration is currently working on a standard for the OBSI method, which is ex-pected
to be adopted by the American Association of State Highway and Transporta-tion
Officials ( AASHTO) as standard AASHTO TP- 76. In Europe the Close Proximity
method ( CPX) is currently used to perform detailed measurements of tire noise emis-sion
from road pavements.
Reliable and accurate noise data is an important factor for efficient implementation
and use of noise reducing pavements by road administrations. The objective of this
current report is to analyze how the temperature affects the results of noise measure-ments
performed according to the OBSI method as it is currently applied by the
UCPRC, through the use of an SRTT test tire. The results are also relevant for the
CPX method with an SRTT applied. The report can also be seen as a contribution to
the ongoing international work on development of standardization of noise measure-ment
methods like the CPX and wayside measurements like the Statistical Pass- by
method ( SPB) etc. An overview of international results is presented as an introduction.
The analysis is based on a unique series of detailed noise measurements performed
on the Caltrans test sections for noise reducing pavement at State Route 138 in the
Mojave Desert in Southern California. The measurements were carried out in the
desert within three consecutive days in the wintertime where the variation of the air
temperature over the day was from to 2 to 22° C. This secures that the main variable
parameter during these measurements is the temperature. A series of other similar
measurement results performed by the UCPRC in the Davis Sacramento area is also
included.
The project has been carried out under the framework of the research technical agree-ment
titled “ Supplementary Studies for the Caltrans Quieter Pavement Research Pro-gram”
between Caltrans and UCPRC as a part of the task: “ Policy documents: guide-lines
for Caltrans policy”. The Danish Road Institute ( DRI- DK) was subcontracted by
UCPRC to work on the project. The work was carried out by a project group with the
following members:
• Hans Bendtsen, Danish Road Institute/ Road Directorate ( DRI- DK) working as a
guest researcher at UCPRC.
• Qing Lu, University of California Pavement Research Center.
• Erwin Kohler, Dynatest Consulting Inc.
12
Erwin Kohler was responsible for the OBSI measurements, collected in the field by
Mark Hannum, of the UCPRC, as part of a Caltrans project “ Third Year Monitoring of
Asphalt Pavement Sections” Partnered Pavement Research Center Strategic Plan Ele-ment
4.19. The data analysis was performed by Qing Lu, UCPRC and the report has
been written by Hans Bendtsen, DRI- DK. Bent Andersen ( DRI- DK) has taken part in
the evaluation and discussion of the results and he has performed a Quality Assess-ment
of the report.
13
Forord
Internationale erfaringer viser, at temperaturen er en faktor som har en vis indflydelse
på resultaterne af målinger af vejtrafikstøj. On Board Sound Intensity ( OBSI) metoden
anvendes ved University of California Pavement Research Center ( UCPRC)
såvel som af andre forskere og konsulenter i USA til at foretage detaljerede målinger
af dækstøjemission fra vejbelægninger. OBSI metoden anvendes tit i støjprojekter
udført for California Department of Transport ( Caltrans). En ekspertgruppe nedsat af
USA's Federal Highway Administration arbejder i øjeblikket på en standard for OBSI
metoden, som forventes at blive vedtaget af American Association of State Highway
and Transportation Officials ( AASHTO) som standard AASHTO TP- 76. I Europa
anvendes Close Proximity metoden ( CPX) til at udføre detaljerede målinger af dæk/
støjemission fra vejbelægninger.
Pålidelige og præcise støjdata er en vigtig faktor for en effektiv implementering og
anvendelse af støjreducerende belægninger i vejforvaltninger. Formålet med denne
rapport er at analysere, hvordan temperaturen påvirker resultaterne af støjmålinger
udført efter OBSI metoden, som i øjeblikket anvendes i UCPRC, med anvendelse
af et SRTT testdæk. Resultaterne er også relevant for CPX- metoden, hvor der anven-des
et SRTT dæk. Rapporten kan også ses som et bidrag til det igangværende interna-tionale
arbejde med udvikling af standardisering af støjmålemetoder som CPX og
den Statistiske Pass By- metode ( SPB) osv. En oversigt over hidtidige internationale
resultater præsenteres som en introduktion.
Analysen er baseret på en unik serie af detaljerede støjmålinger udført på Caltrans test
sektioner med støjreducerende belægninger på vej LA138 i Mojave ørkenen i det syd-lige
Californien. Målingerne blev foretaget i ørkenen inden for tre dage i vintermåne-derne,
hvor variationen af lufttemperaturen hen over dagen var fra 2 til 22 ° C. Dette
sikrer, at den vigtigste variabel i løbet af disse målinger er temperaturen. En række an-dre
resultater af lignende målinger udført af UCPRC i Davis Sacramento området er
også medtaget.
Projektet er gennemført inden for rammerne af en aftale med titlen ” Supplerende Un-dersøgelser
for Caltrans ” Quieter Pavement Research Program”” mellem Caltrans og
UCPRC som en del af opgaven: ” Policy documents: guidelines for Caltrans policy”.
Vejdirektoratet/ Vejteknisk Institut har været kontraheret af UCPRC til at udføre en del
af arbejdet. Arbejdet er udført af en projektgruppe med deltagelse af følgende perso-ner:
• Hans Bendtsen, Vejdirektoratet/ Vejteknisk Institut ( DRI- DK) der arbejdede som
gæsteforsker på UCPRC fra august 2008 til august 2009.
• Qing Lu, University of California Pavement Research Center.
• Erwin Kohler, Dynatest Consulting Inc.
14
Erwin Kohler var ansvarlig for OBSI målingerne udført af Mark Hannum fra UCPRC,
som en del af en Caltrans projektet " Third Year Monitoring of Asphalt
Pavement Sections". Dataanalysen blev udført af Qing Lu, UCPRC og rapporten er
skrevet af Hans Bendtsen, DRI- DK. Bent Andersen DRI- DK har deltaget i evaluering
og diskussion af resultaterne, og han har kvalitetssikret rapporten.
15
1. Introduction and existing knowledge
Different methods are used to measure noise emission caused by road traffic passing
over a specific pavement. Noise measurements are often carried out with the objective
of measuring the noise properties of a specific road surface. High levels of accuracy
are needed in such measurements as the difference in noise emission between different
pavements is often quite small. From international experience it is known that tem-perature
influences the noise generated by road traffic. There is therefore a need for
knowledge on the influence of temperature in relation to the different measurement
methods used. In this report the main focus is on the On Board Sound Intensity
( OBSI) method used in California and in other U. S. states. The results will also have
relevance for other noise measurement methods ( see Section 1.2) like the Close Prox-imity
method ( CPX) with an SRTT applied.
In this project, the temperature will generally be given in degrees Celsius (° C) and
when relevant also the temperature in Fahrenheit (° F) will be given. The box below
states the transformations between these two units of temperature.
Temperature correction: ( TFahrenheit - 32)* 5/ 9 = TCelsius
Temperature coefficient correction: cFahrenheit * 1.8 = cCelsius ( see Section 1.2)
All the noise levels presented in this report are A- weighted. The unit “ dB” is used in
this report and it is equal to what is often denoted “ dB( A)” and “ dBA”.
1.1 The “ Tyre/ Road Noise Reference Book”
The Tyre/ Road Noise Reference Book by Sandberg and Ejsmont from 2002 [ 6] in-cludes
a summary of international status of the current knowledge at that time on the
influence of temperature on tire/ road noise generation. The general knowledge was
that the tire/ road noise from automobile tires is influenced by about - 1 dB per 10 ° C
temperature increase. It is stated that the current problems were:
• that the mechanisms by which the noise generation are influenced by temperature
were not properly understood,
• that the measured effects of temperature have varied greatly,
• that it had been difficult to see any general rule that could be practically applied.
There are two major friction/ adhesion related tire pavement noise generating mecha-nisms
described as hypotheses by Sandberg and Ejsmont in [ 6] ( see Figure 1.1):
• “ The first mechanism is the “ stick- slip” mechanism in which tangential stresses in
the rubber- road interface are built up and released. This causes a tangential vibra-tion
that might be called “ scrubbing”. When the surface is not perfectly flat, the
vibrations affected by this process may have both radial and tangential compo-nents.”
16
• “ The “ stick- slip” mechanism will give increased noise emission when friction is
increased, in particular at high frequencies, and in particular for tires with small
tread pattern depth.”
• “ The second one is a “ stick- snap” mechanism due to adhesive bonds between rub-ber
and road which are broken at a certain level when rubber is “ pulled away”
from the road contact. This may cause a combination of radial and tangential vi-brations,
but the sudden release of a rubber block from the surface may also cause
a transient air- flow through the opening slit.”
• “ The stick- snap mechanism will give increased noise when the attraction force be-tween
rubber and the road surface is increased. This is not necessarily closely re-lated
to the tangential friction characteristics important for the stick- slip, but more
related to having a very close and unbroken rubber- surface contact. An extremely
smooth surface might provide such conditions. However, it depends largely also on
material properties; i. e., whether and to what extent the materials are hydrophobic
( have high attraction to each other) or hydrophilic ( have low attraction).”
• “ An increased microstructure will normally give increased friction, and thus in-creased
stick- slip motion amplitudes, but it may give decreased adhesion bond
strengths, which will reduce stick- snap effects.”
Figure 1.1. Illustration of the “ stick- slip” and the adhesion “ stick- snap” tire- road noise generating
mechanisms [ 6] ( used with permission from Ulf Sandberg, VTI).
It could be anticipated that the adhesion “ stick- snap” tire- road noise generating
mechanisms will mainly lead to increased noise levels at the “ back” end of the tire
where the rubber blocks “ leave” the pavement surface. For “ stick- slip”, the scrubbing
of the rubber on the pavement will occur at both the back and the front of the tire. In
the OBSI method, noise is measured both in front and behind the tire. It could be ana-lyzed
if there is a systematic difference of these two noise levels and if such a differ-ence
varies with temperature.
On the background of available data the following general trends are presented in [ 6]:
1. Tire temperature is not very useful for considering a correlation between noise
and temperature.
17
2. There seems to be no clear benefit in using road temperature instead of air tem-perature
or vice versa as a temperature descriptor.
3. The effect of speed on the noise temperature relation is inconsistent.
4. There is a big range in temperature coefficients from - 0.03 to - 0.20 dB/° C for
different passenger car tires.
5. For truck tires the temperature coefficient is much lower.
6. There is a big range in temperature coefficients from - 0.03 to - 0.20 dB/° C for
different pavement types.
7. The temperature coefficients are clearly frequency dependent.
8. The tangential stiffness of an asphalt surface may be influenced by temperature
which could potentially influence the noise generation from the “ stick- slip”
process ( see Figure 1.1) where the rubber tread blocks motions relative to the
road surface causing tangential tire vibrations presumably over 1000- 2000 Hz.
9. It has been suggested to develop a model for the noise- temperature relation as a
function of the elastic modulus of the tire tread compound or the tread hardness.
Work on the subject is under way within the International Organization for Standardi-zation
( ISO), but is currently not finalized as of the end of 2008.
1.2 Temperature and different noise measurement methods
Two different types of noise measurement methods are commonly used:
1. The “ close to source” methods where the noise is measured near the
tire/ pavement interface:
a. The On Board Sound Intensity method [ 1] where the sound intensity is
measured by microphone probes placed very close to the contact point be-tween
the tire and the road surface ( see Chapter 3). The measurement
equipment is mounted on a passenger car. Here the temperatures of the ac-tual
test tire as well as the pavement temperature are relevant parameters.
This method is currently used in California and other U. S. states. Devel-opment
of a standard for the OBSI method is ongoing in the U. S.
b. In the Close Proximity method ( CPX) [ 2], sound pressure levels are meas-ured
by microphones placed very close to the contact point between the tire
and the road surface. The measurement equipment is mounted either on a
trailer or a passenger car. Here the temperatures of the actual test tire/ tires
as well as of the pavement are relevant parameters. This method is com-monly
used in Europe. Work is ongoing on finalizing an ISO standard for
the CPX method.
18
2. The “ roadside methods” where noise is measured at the road side:
a. In the Statistical Pass- by method ( SPB) [ 3], noise is measured from ran-domly
chosen single vehicles driving at constant speed at a distance of 7.5
m between the microphone and the center line of the lane and at a height of
1.2 m. Here the average temperatures of the tires of all the selected vehi-cles
as well as of the pavement are relevant parameters.
b. In the Controlled Pass- by method ( CPB), noise is measured from one or a
few selected vehicles at the same microphone position as the SPB method.
Here the temperatures of the tires of the one or few selected vehicles as
well as of the pavement are relevant parameters.
c. LAeq measurements where the noise from the vehicles passing by is meas-ured
over a longer period.
From the above it can be seen that depending on the method used either the average
tire temperature of the vehicles included in the measurements or the temperature of the
test tire/ tires used are important together with the pavement temperature.
None of general specifications for these methods today include procedures for tem-perature
correction. But some of the organizations in Europe using these noise meas-urement
methods have developed their own practice for making temperature correc-tions.
The Danish Road Institute ( DRI- DK) uses for example the air temperature and
applies the following corrections to SPB measurements with a reference air tempera-ture
of 20 ° C ( 68 ° F) [ 5]:
T corr; P = 0.05 · ( T measured – 20) ; Passenger cars
T corr; H = 0.03 · ( T measured – 20) ; Heavy vehicles
The air temperature is recorded approximately every 20 minutes. These temperature
corrections are based on recommendations in a publication from the Dutch organiza-tion
CROW [ 4]. For CPX measurements, DRI- DK uses the same temperature correc-tion
as for passenger cars in the SPB method.
19
Figure 1.2. The pavement temperature as a function of the air temperature ( data from the SPB
measurement series carried out in different European countries), [ 8].
It can be discussed which temperature is the one that is important for the noise genera-tion.
There are three main possibilities:
1. The air temperature.
2. The temperature of the pavement surface.
3. The temperature of the tire/ tires.
The pavement is heated up by the ambient air and radiation from the sun. Figure 1.2
and 1.3 shows different series of simultaneous measurements of air and pavement sur-face
temperature performed in Europe and in California.
Figure 1.2 shows the relation between the pavement surface temperature and the air
temperature based on data from the SPB measurement series carried out in different
European countries [ 8]. There is a rather good linear correlation ( R2 = 0.83) between
pavement and air temperature. The pavement surface temperature was on average 10
° C higher than the air temperature when the air temperature was 30 ° C and the pave-ment
temperature was on the average a little lower ( 3 ° C) than the air temperature
when the air temperature was 5 ° C. In general, for a given air temperature there was a
± 5 ° C variation in pavement temperature.
20
- 5
0
5
10
15
20
25
30
35
- 5 0 5 10 15 20 25
Air Temperature ( C)
Pavement Temperature ( C)
Figure 1.3. Simultaneously measured pavement and air temperature at some Caltrans test sections
on highway LA138 in Mojave desert in Southern California. The measurements were carried out in
the wintertime.
Figure 1.3 shows the results of simultaneously measured pavement and air tempera-ture
at some Caltrans test sections on highway LA138 in Mojave Desert in the winter
period. Until the air reaches around 20 ° C the pavements are colder than the air then
the pavements starts to heat up faster than the air.
The tire road noise is generated by vibrations in the tires caused by the roughness of
the pavement surface as well as by air pumping and other mechanisms ( see section
1.1). If the contact between the rubber blocks of the tread pattern of the tire becomes
soft or elastic, the noise will be reduced. The temperature might influence this in two
different ways:
1. By a more elastic pavement surface caused by higher temperatures.
2. By a softer rubber in the tread pattern of the tire caused by a higher temperature.
In the European SILVIA project [ 9] the influence of pavement elasticity on noise gen-eration
was analyzed [ 10]. It was concluded that the stiffness of present pavements is
much larger than the tire stiffness and that a reduction of the noise is only possible if
the pavement stiffness is in the same order of magnitude as the tire stiffness ( pavement
stiffness/ tire stiffness < 10).
21
This is not at all the case with normal asphalt and concrete pavements, and an increase
in temperature cannot reduce the pavement stiffness to a stiffness which is in the same
order of magnitude as the stiffness of a rubber tire. This will only be possible if alter-native
materials like rubber are used for pavement construction instead of rock aggre-gate.
On this background it can be concluded that it is not a change in the pavement tem-perature
that affects the noise properties in relation to elasticity of a “ normal” pave-ment.
This means the temperature affects the noise properties of the tires. It can be an-ticipated
that when the tire gets warmer, the rubber becomes softer and this influ-ences/
reduces the vibration generated noise and possibly the “ stick- slip” process.
Therefore the tire temperature is a relevant parameter for estimating the temperature
effect on the tire- pavement rolling noise generation.
The temperature of a tire must be defined by the ambient air temperature as well as
by the heat generated in the tire when the tire is deformed while rolling over the
pavement. The air presumably heats/ cools the tire until an equilibrium tire temperature
is reached. The tires only touch the pavement at a small contact area during a short
time, and therefore the pavement temperature cannot be the most significant factor for
the tire temperature.
If tire temperature measurements are not available, the air temperature might be re-garded
a better indicator of the tire temperature than the pavement temperature even
though this can be discussed. In this project, noise will be analyzed both in relation to
air as well as to pavement temperature. Tire temperatures have not been available.
In the following, a series of international results for the last ten years will be pre-sented.
1.3 Results from the German- Dutch Sperenberg project
A closed military airport ( Sperenberg) near Berlin in Germany has been turned into a
test facility for different pavement types. A total of 46 different pavements have been
constructed. Noise and other pavement properties have been measured intensively [ 7].
A survey of the influence of temperature has also been performed at Sperenberg by
application of the Controlled Pass- by noise measurement method for two passenger
cars with eight different tires and one truck with four different tires. Six different
pavements were included in the measurement series. The coast- by noise without the
engine running has been measured using a roadside microphone position placed at a
height of 1.2 m above the pavement and 7.5 meters from the centre line of the vehicle
passing by. The measurements were carried out in an air temperature range between 0
and 35 ° C. Some main results are presented in the following.
22
Figure 1.4. A Mercedes passenger car on the Sperenberg pavement test site with 46 different
pavements on a closed military airport near Berlin in Germany.
Noise measurements have been performed for two passenger cars. A Mercedes with
eight different tires ( called M1 to M8, typical dimension 195/ 65 R15) and a VW Polo
also with eight different tires ( called W1 to W8, typically 175/ 70 R13). A truck with
four different tires was also included ( called T1 to T4, 315/ 80 R22.5). For each cate-gory
tire No. 1 is not a normal tire, but a slick tire. The following linear regression
model has been used to describe the noise level ( LA, max):
LA, max = a + b ( 10 log( v/ vo)) + c ( Tair- 20)
Where v is the vehicle speed and vo is a reference speed of 80 km/ h for passenger cars
and 70 km/ h for trucks. c is the regression coefficient for the air temperature in dB/° C.
A c- value of - 0.05 dB/° C means that the noise decreases 0.5 dB when the air tempera-ture
increases 10 ° C.
Some main results from measurements on a dense asphalt concrete with a maximum
aggregate size of about 8 mm and a porous pavement also with a maximum aggregate
size of 8 mm are shown in Figure 1.5. It can be seen that there is quite a big variation
of c for the different tires on the same car. For example, c varies for dense asphalt
pavements between - 0.037 and - 0.129 dB/° C for the different tires on the Mercedes
passenger car. The picture is the same for the different tires on the VW- Polo (- 0.062 to
- 0.131 dB/° C). The air temperature coefficient c is smaller for the truck tires (- 0.039 to
- 0.055 dB/° C). These measurements indicate that for noise measurement methods like
OBSI and CPX as well as the CPB that uses specific tires it is necessary to apply spe-cial
temperature corrections that are related to the specific tires used, whereas for the
Statistical Pass- By method an average temperature correction seems relevant for each
vehicle category.
23
- 0,14
- 0,12
- 0,1
- 0,08
- 0,06
- 0,04
- 0,02
0
M1 M2 M3 M4 M5 M6 M7 M8 W1 W2 W3 W4 W5 W6 W7 W8 T1 T2 T3 T4
Air temperature coefficient c [ dB/ oC]
Dense Asphalt
Porous Asphalt
Mercedes VW- Polo Truck
Figure 1.5. Air temperature coefficients in dB/° C for pass- by noise measurements of two passenger
cars with 8 different tires and a truck with 4 different tires on dense asphalts and porous pave-ments
[ 7].
Figure 1.6 shows the same type of data for a cement concrete pavement again com-pared
to the dense asphalt concrete also presented in Figure 1.5. The temperature coef-ficients
c are smaller on the cement concrete pavement than on the asphalt pavement.
Table 1.1. The temperature coefficient c in dB/° C for the 3 vehicles averaged over all the tires used
on these vehicles for the three pavement types [ 7].
Pavement type Dense asphalt Porous asphalt Cement concrete
Mercedes car ( 8 different tires) - 0,091 - 0,073 - 0,044
VW- Polo car ( 8 different tires) - 0,089 - 0,049 - 0,042
Truck ( 4 different tires) - 0,048 - 0,020 0,001
Table 1.1 shows the air temperature coefficient c for the three vehicles averaged over
all the tires used on these vehicles for the three pavement types. From these data it
seems that the temperature effect on tire- road noise is around twice as high for pas-senger
cars than for trucks. It also seems that the temperature coefficient depends on
the pavement type. The tires at the dense asphalt have the highest temperature coeffi-cient,
around twice the coefficient for the porous asphalt pavement. The cement con-crete
pavement has the lowest coefficient but the variation for the concrete pavement
is high.
24
- 0,14
- 0,12
- 0,1
- 0,08
- 0,06
- 0,04
- 0,02
0
0,02
0,04
0,06
0,08
M1 M2 M3 M4 M5 M6 M7 M8 W1 W2 W3 W4 W5 W6 W7 W8 T1 T2 T3 T4
Air temperature coefficient c [ dB/ oC]
Dense Asphalt
Concrete
Figure 1.6. Air temperature coefficients in dB/° C for pass- by noise measurements of two passenger
cars with eight different tires and a truck with four different tires on dense asphalt and cement
concrete pavements [ 7].
The frequency dependency of the air temperature coefficient is shown for dense as-phalt
and cement concrete pavements in Figure 1.7 averaged for the normal passenger
car tires included in the measurements at Sperenberg. At low frequencies, the tempera-ture
coefficient is quite low. At frequencies over 630 to 1000 Hz, the temperature co-efficient
is around - 0.12 dB/° C for asphalt pavement and - 0.10 dB/° C for cement con-crete
pavement.
- 0,14
- 0,12
- 0,1
- 0,08
- 0,06
- 0,04
- 0,02
0
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000
5000
Third octave band frequency [ Hz]
Air temperature coefficient [ dB/ oC]
Dense asphalt
Concrete
Figure 1.7. Average spectral temperature coefficient in dB/° C at different frequencies for selected
passenger car tires at dense asphalt and cement concrete pavements [ 7].
25
Some general conclusions from the Sperenberg temperature study [ 7] are:
• With increasing temperature, the noise is decreased.
• This temperature effect is not significantly dependent on the vehicle speed
( measured in the range from 50 to 110 km/ h.
• The temperature is most dominant in the frequency range from 630/ 1000 to
5000 Hz.
• The temperature effect varies a lot for different tires.
• The temperature effect varies for the three different asphalt pavement types but
the trend is not very consistent.
• The temperature effect varies for the three different pavement types but the trend
is not very consistent.
• The temperature effect is higher on asphalt pavements than on cement concrete
surfaces.
• The temperature effect for passenger car tires is approximately twice the effect
for truck tires.
1.4 Semi- generic temperature correction method
At the Inter. Noise conference in 2004 in Prague, Ulf Sandberg suggested a method for
temperature corrections [ 18]. A “ semi- generic” correction method has been devel-oped,
where a correction factor is specified separately for each major group of tires
and each major group of road surfaces. The method is developed on the background of
the empirical data measured in the Sperenberg experiment [ 7] ( see Section 1.3) as well
a by expert judgments. The air temperature is used as the temperature indicator. The
input for the method are pavement characteristics like texture expressed as MPD and
air voids. The selected air temperature correction coefficients are predicted as average
values for a series of different tires ( basically the 20 tires included in the Sperenberg
experiment). Therefore this method is developed to be used for correction of noise
measurements including a large series of different vehicles with different tires on the
same road surface like typically the SPB method as well as LAeq measurements over
longer periods ( see Section 1.2). The method is not suitable to be used directly for
OBSI or CPX measurements where only one or a few specific tires are used.
The suggested air temperature coefficients for passenger cars can be seen in Table 1.2.
For trucks the values have to be divided by 2 [ 18]. Due to lack of data, frequency-dependent
temperature coefficients were not suggested [ 18].
26
Table 1.2. Proposed air temperature coefficients in dB/° C for passenger cars for various types of
road surfaces [ 18]. Values in parenthesis are for uncommon surfaces for which there were no avail-able
measurement data.
Pavement
type
Texture Dense
Air void 0- 8 %
Open graded
Air void 8- 15 %
Porous
Air void > 16 %
Smooth
MPD < 0.7 mm
- 0.10 - 0.08 (- 0.06)
Medium
0.7 < MPD < 1.4 mm
- 0.06 - 0.06 - 0.05
Asphalt
Concrete
Rough
MPD > 1.4 mm
- 0.12 - 0.06 - 0.04
Smooth
MPD < 0.7 mm
- 0.05 (- 0.04) (- 0.04)
Cement
Concrete
Medium to rough
MPD > 0.7 mm
- 0.09 (- 0.04) - 0.03
All other
surfaces
Any (- 0.06) (- 0.05) (- 0.04)
1.5 Previous American investigations
At the TRB ADC40 noise and vibration meeting in Key West in July 2008, three dif-ferent
series of measurement data on the effect of temperature on OBSI measurements
were reported.
Paul R. Donavan and Dana M. Lodico presented measurements on a Dense Asphalt
Concrete ( DGAC) and a Portland Cement Concrete ( PCC) pavement performed with
a SRTT ( see Figure 1.8) as well as a Dunlop SP Winter Sports tire [ 15 and 22]. Noise
measurements were performed at different air temperatures ranging from 30
to 40 ° C ( 86 to 104° F).
27
Figure 1.8. A Standard Reference Test Tire ( SRTT) to the left ( photo Bruce Rymer, Caltrans) and the
Dunlop SP Winter Sports tire to the right ( photo Paul A. Donavan, Illingworth & Rodkin, Inc).
The pavement temperatures were ranging from 35 to 61 ° C ( 95 to 142 ° F). For the
Dunlop tire, the air temperature coefficient c was nearly the same for the DGAC
pavement and for the PCC pavement, respectively - 0.100 and - 0.086 dB/° C. No clear
correlations were found for the SRTT at this temperature range, but the data indicate
that the air temperature coefficient c is around - 0.024 dB/° C for DGAC pavement and
- 0.027 for the PCC pavement.
Judith Rochat and Aron Hastings [ 16] presented two series of measurements. The first
was performed in Arizona on a transversely tined Portland Cement Concrete ( PCC)
pavement, and the second on two Asphalt Rubber Friction Courses ( ARFC) which
were new and one year old respectively. The noise measurement method used was
roadside LAeq measurements over 5 minute periods on the actual traffic passing the
measurement position. This means that these measurements included different vehi-cles
and different vehicle categories with many different tires contrary to OBSI meas-urements.
Noise measurements were performed at different air temperatures ranging
from 29 to 39 ° C ( 85 to 102 ° F). The pavement temperatures were ranging from 29 to
51 ° C ( 84 to 124 ° F). The results can be seen in Figure 1.10. The pavement tempera-ture
coefficient c varied between - 0.018 and - 0.072dB/° C. For the air temperature co-efficient
c the variation was between - 0.043 and - 0.160 dB/° C. The average air tem-perature
coefficient for ARFC was - 0.064 dB/° C and for the PCC it was - 0.130 dB/° C.
28
Figure 1.9. Road with Asphalt Rubber Friction Course ( ARFC) in Phoenix Arizona.
- 0,17
- 0,15
- 0,13
- 0,11
- 0,09
- 0,07
- 0,05
- 0,03
- 0,01
0,01
PCC ARFC new ARFC 1 year
Temperature coefficient c [ dB/ oC]
Site 1
Site 2
Pavement temperature in Celcius
- 0,17
- 0,15
- 0,13
- 0,11
- 0,09
- 0,07
- 0,05
- 0,03
- 0,01
0,01
PCC ARFC new ARFC 1 year
Temperature coefficient c [ dB/ oC]
Site 1
Site 2
Air temperature in Celcius
Figure 1.10. Temperature coefficients in dB/° C for pavement temperature respectively air tempera-ture
measured in Arizona using roadside L Aeq measurements over 5 minute periods on the actual
traffic [ 16].
29
Figure 1.11. The roadside SPB noise measurement setup at LA138 test road in the Mojave Desert
( Photo Judith Rochat, VOLPE).
- 0,170
- 0,150
- 0,130
- 0,110
- 0,090
- 0,070
- 0,050
- 0,030
- 0,010
0,010
DGAC OGAC 75
Temperature coefficient c [ dB/ oC]
Cars
Trucks
Pavement temperature in Celcius
- 0,170
- 0,150
- 0,130
- 0,110
- 0,090
- 0,070
- 0,050
- 0,030
- 0,010
0,010
DGAC OGAC 75
Temperature coefficient c [ dB/ oC]
Cars
Trucks
Air temperature in Celcius
Figure 1.12. Temperature coefficients in dB/° C for pavement temperature respectively air tempera-ture
measured on LA138 using the roadside SPB method [ 16].
30
Results from a series of roadside SPB noise measurements are also presented in [ 16].
The measurements were performed on a Dense Graded ( DGAC) and an Open Graded
( OGAC 75) asphalt concrete pavement on the LA138 test road in the Mojave Desert
( see Chapter 2). These measurements include different vehicles with many different
tires contrary to OBSI measurements. New OBSI measurements on these pavements
at different temperatures will be presented in Chapter 4. Noise measurements were
performed at different air temperatures ranging from 8 to 32 ° C ( 47 to 90 ° F). The
pavement temperatures were ranging from 7 to 49 ° C ( 45 to 121 ° F). The results can
be seen in Figure 1.12. For the air temperature coefficient c, the variation was between
- 0.022 and - 0.049 dB/° C for passenger cars, whereas it was slightly positive for trucks
( 0,009 to 0,016 dB/° C). The temperature coefficients in dB/° C measured at LA138
were generally significantly lower than the coefficients measured in Arizona.
1.6 Results from a French experiment
In an article in Applied Acoustics from 2007, Fabienne Anfosso- Lédée and Yves
Pichaud present results from a French experiment [ 19] carried out on the test tracks of
the National French Road Laboratory ( LCPC) in Nantes ( see Figure 1.13). Two differ-ent
Michelin summer tires were used ( a more “ noisy” Tire A and a “ low noise” Tire B
( see Figure 1.14)). The rubber hardness of the A tire was 76.3 shore A and of the B
tire 79.5 shore A [ 21]. The noise measurements have been performed on seven differ-ent
dense and open graded pavements ( see Table 1.3) including asphalt and cement
concrete as well as surface dressings.
Figure 1.13. The LCPC test tracks in Nantes.
One test vehicle driving at constant speed was used. The noise measurements were
performed by the roadside Controlled Pass- By ( CPB) method with a microphone
placed at a height of 1.2 m and a distance of 7.5 m from the centre line of the test ve-hicle.
The results are presented for a speed of 90 km/ h ( 56 mph). To cover a large
temperature range, the measurements were performed at different seasons over the
year 2000 to 2001 covering a range in air temperatures from 0 to 30 ° C. Pavement, air
and tire temperatures were measured.
31
The following temperature relations in degrees Celsius were found:
Troad = 1.7Tair – 4.5 [° C]
Ttire = 1.05Tair + 15.8 [° C]
Table 1.3. The seven pavements included in the French experiment [ 19].
Name Type Maximum aggregate
size in mm
Mean Profile Depth
( MPD) in mm
DGAC Dense asphalt concrete 10 0.86
PAC Porous asphalt concrete 10 1.67
OGAC
Very thin open graded
asphalt concrete
10 1.49
SD rough
Rough epoxy bound
surface dressing
10 4.3
SD fine
Thin and smooth epoxy
bound surface dressing
1.5 0.70
PCC burlap
Burlap textured cement
concrete
- 0.80
PPCC Porous cement concrete - 1.14
Figure 1.14. The two Michelin summer tires used in the French experiment. The “ noisy” Tire A to
the left and “ low noise” Tire B to the right ( photo Fabienne Anfosso- Lédée, LCPC).
32
- 0,170
- 0,150
- 0,130
- 0,110
- 0,090
- 0,070
- 0,050
- 0,030
- 0,010
0,010
DGAC PAC OGAC SD rough SD fine PCC burlap PPCC
Temperature Coefficient c [ dB/ oC]
Tire A
Tire B
Figure 1.15. Air temperature coefficients in dB/° C measured at the LCPC test tracks using the road-side
CPB method [ 19].
Air temperature coefficients can be seen in Figure 1.15. The results from the two
tires are practically the same and the coefficients vary between - 0.02 and - 0.13 dB/° C.
The results are grouped for three pavement types: dense and open asphalt concrete as
well as cement concrete ( see Table 1.4). The air temperature coefficient is - 0.10 dB/° C
for the dense pavements and - 0.06 dB/° C for the open graded ( and porous) pavements.
For both the porous and dense cement concrete the coefficient is - 0.03 dB/° C.
Table 1.4. Average temperature coefficients from the French study grouped for three pavement
types [ 19].
Pavement type Average Air
temperature
coefficients
Average
Pavement
temperature
coefficients
Average Tire
temperature
coefficients
Range in MPD
Dense asphalt concrete - 0.10 dB/° C - 0.06 dB/° C - 0.09 dB/° C 0.70 – 0.86 mm
Open asphalt concrete - 0.06 dB/° C - 0.04 dB/° C - 0.05 dB/° C 1.49 – 4.3 mm
Cement concrete - 0.03 dB/° C - 0.02 dB/° C - 0.03 dB/° C 0.80 – 1.14 mm
Spectral analysis [ 19] showed that in general the noise in the low frequencies ( below
500 Hz) and in the high frequencies ( from 1600 to 5000 Hz) seems to be affected by
temperature. For higher frequencies, the noise levels are 2- 3 dB lower for higher tem-peratures
than for lower temperatures.
1.7 The European Union Tire Noise Directive
The European Union has a directive that regulates the noise emission from new tires
sold in the Union [ 20]. In Annex 5 of this directive there is a description of the test
procedures for measuring tire noise emission. The measurement method is a coast by
method where the noise is measured at the roadside ( distance 7.5 m and height 1.2 m)
while test vehicles equipped with the tires to be tested are driving on a specified dense
asphalt concrete surface with a maximum aggregate size of 8 mm.
33
A reference speed of 80 km/ h is used for tires for passenger cars and vans/ small trucks
where as the reference speed for tires for heavy vehicles is 70 km/ h. Measurements of
air as well as test pavement temperature are mandatory. Measurements shall not be
made when the air temperature is below 5 ° C or above 40 ° C or when the test pave-ment
temperature is below 5 ° C or above 50 ° C. A reference speed of 80 km/ h is used
for tires for passenger cars and vans/ small trucks where as the reference speed for tires
for heavy vehicles is 70 km/ h. Prior to testing, tires shall be warmed up by running
under test conditions.
The final results are normalised to a test pavement reference temperature of 20 ° C us-ing
the following pavement temperature correction factors:
• Passenger cars ( called type C1) – 0.03 dB/° C when the pavement temperature is
over 20° C and – 0.06 dB/° C when the pavement temperature is under 20 ° C.
• Vans/ small trucks ( called type C2) – 0.02 dB/° C.
• Heavy vehicles ( called type C3) no temperature correction.
1.8 The challenge
A hypothesis could be that the “ stick- snap” and the adhesion “ stick- slip” mechanisms
might be influenced by temperature ( see Section 1.1.). These mechanisms are thought
to lead to increased noise levels at higher frequencies above 1000 to 2000 Hz.
The measurement series both from Sperenberg ( see Section 1.3) and France
( see Section 1.6) showed that high frequency noise is reduced with increasing tem-peratures.
These two tire- road noise generating mechanisms might mainly lead to
increased noise levels at the “ back” end of the tire where the rubber blocks “ leaves”
the pavement surface. In OBSI and the CPX methods, noise is measured both in front
and behind the tire. It could be analyzed if there is a systematic difference between
these two noise levels and if such a difference varies with temperature. Such analyses
have not been carried out in this current project.
The Sperenberg and the French results also indicate that both the properties of the tire
as well as the pavement type have an influence on the temperature coefficient. As
regards the tire influence, a hypothesis can be that increased temperature makes the
rubber compound softer and this reduces the vibration generated noise from the tires.
If this was the case, the temperature influence should occur in the lower frequencies,
but as already mentioned the Sperenberg results show increased noise levels at higher
frequencies above 1000 to 2000 Hz.
As regards pavement influence one of the conclusions from the SILVIA project was
that ( see Section 1.1) the stiffness of present pavements is much larger than the tire
stiffness, and that a reduction of the noise is only possible if the pavement stiffness is
in the same order of magnitude as the tire stiffness. Pavements are normally not as
stiff in the same order of magnitude as tire rubber even under very warm weather con-ditions,
which indicates that other noise- related properties than the pavement stiffness
might be influenced by increased temperature. These could be the “ stick- snap” and the
“ stick- slip” mechanisms.
34
In this report the main objective is to investigate the influence of temperature on the
On Board Sound Intensity method currently applied in California through the use of
an SRTT test tire. The results are also relevant for the CPX method with an SRTT ap-plied.
This is done by analyzing two sets of measurement data:
1. A series of detailed OBSI noise measurements performed by UCPRC on the
Caltrans test sections for noise reducing pavements at highway LA138 in the
Mojave Desert in Southern California [ 11]. The measurements are carried out in
the desert in the wintertime where the variation of the air temperature over the
day was from to 2 to 22 ° C and pavement temperatures from to - 1 to 33 ° C.
Here the noise has been measured on the same day or within a few consecutive
days with the same equipment, by the same operator, and on the same pave-ments,
at low morning, medium midday and high afternoon temperatures. This
ensures that the only main variable parameter during these measurements is the
temperature. For these measurements, a Standard Reference Test Tire ( SRTT)
was used.
2. Another measurement series was performed by UCPRC in California as part of
a large project on pavement noise [ 12]. For these measurements, a Goodyear
Aquatred tire was used, which was the old former standard test tire for OBSI.
The objective was to measure the noise at three different temperatures on three
different types of pavements. The variation of the pavement temperature over
the day was from to 11 to 35 ° C. Here the noise was measured on the same day
with the same equipment, by the same operator, and at the same measurement
positions at three different temperatures.
The objective of the two measurement series was to perform measurements where the
only variable was the temperature and where the following factors were constant:
• Same measurement tire.
• Same inflation and rubber hardness ( at a reference temperature) of the measure-ment
tire.
• No changes in age, tear, and wear of the measurement tire.
• Same acoustical measurement equipment.
• Measurement tire mounted on the same car.
• Same measurement operator.
• No changes in wear and tear of pavements.
35
2. The test sections
The pavements for the two sets of test sections included in this project are presented
in the following.
2.1 The LA138 pavements
The LA138 test sections were constructed on State Highway 138 in the Mojave Desert
west of Lancaster in 2001. The purpose was to develop and test different types of
noise reducing pavements [ 11]. A total of 5 different pavements were constructed
including a Dense Graded Asphalt Concrete ( DGAC) used as a reference. The OBSI
measurements were performed both in the eastbound and the westbound directions
in February 2008 when the pavements were 8 years old. The DGAC was for practical
reasons only measured in one direction. Therefore a total of 9 datasets are included in
this survey.
Figure 2.1. The LA138 test road on Highway 138 in the Mojave Desert.
The following pavements were constructed on the test road ( see Table 2.1):
• A Dense Graded Asphalt Concrete ( DGAC) with a specified thickness of
30 mm used as a noise reference pavement.
• An Open Graded Asphalt Concrete ( OGAC 30) with a specified thickness of
30 mm.
• An Open Graded Asphalt Concrete ( OGAC 75) with a specified thickness of
75 mm.
• An Open Graded Asphalt Concrete with rubber powder added to the bitumen
( RAC- O) and a specified thickness of 30 mm.
• A Bonded Wearing Course ( BWC). A propriety product used in California.
36
Table 2.1. Close up pictures of the LA138 test pavements and results of a visual inspection per-formed
in October 2008. The size of the black and white squares at the photos is 10 mm times 10
mm.
Photo October 2008 Visual inspection October 2008
S1 DGAC The pavement is generally in a good condition. Very
small signs of a little raveling are not considered to affect the noise
generation. There are transversal cracks at a width of 2- 5 mm at the
whole width of the lane at approximately each 5 m. Not possible to
hear increased noise at the roadside when tires were passing the
cracks.
S2 OGAC 75 The pavement has an open “ negative” surface struc-ture.
When water was poured on the pavement it did not signifi-cantly
penetrate down into the surface structure ( not porous). The
pavement is generally in a good condition. Very small signs of a lit-tle
raveling are not considered to affect the noise generation. Trans-versal
cracks at a width of 2- 5 mm at the whole width of the lane at
approximately each 5 m. Not possible to hear increased noise when
tires were passing the cracks.
S3 OGAC 30 The pavement has an open “ negative” surface struc-ture.
When water was poured on the pavement it did not penetrate
down into the surface structure ( not porous). The pavement is gen-erally
in a good condition. Small signs of a little raveling are not
considered to affect the noise generation. Transversal cracks at a
width of 2- 5 mm at the whole width of the lane at approximately
each 5 m. Not possible to hear increased noise when tires were pass-ing
the cracks.
S4 RAC- O The pavement has an open “ negative” surface struc-ture.
When water was poured on the pavement it did not penetrate
down into the surface structure ( not porous). Small signs of raveling
are not considered to affect the noise generation. Some longitudinal
cracking. Transversal cracks at a width of 2- 5 mm at the whole
width of the lane at approximately each 5 m. Not possible to hear
increased noise when tires were passing the cracks.
S5 BWC This pavement seems to have the roughest surface struc-ture
of the five pavements. The pavement is generally in a good
condition. Practically no raveling. Transversal cracks at a width of
up to 10 mm at the whole width of the lane at approximately each
5 m. Possible to hear slightly increased noise when tires were pass-ing
the cracks.
37
The five pavements all had a maximum aggregate size of 12.5 mm. The three open
graded pavements had a built- in air void content of around 10 % ( measured on drill
cores when they were six years old).
Table 2.2. Data on the LA138 test pavements. The air void is measured on drill cores in 2007.
Site No. Pavement type Maximum
aggregate size
Specified
thickness
Air void
S1 DGAC 12.5 mm 30 mm 7.0 %
S2 OGAC 75 12.5 mm 75 mm 10.6 %
S3 OGAC 30 12.5 mm 30 mm 10.3 %
S4 RAC- O 12.5 mm 30 mm 10.7 %
S5 BWC 12.5 mm 30 mm 5.0 %
To describe the pavements in relation to noise, the result of road side Statistical Pass-
By measurements for passenger cars when the pavements were 16 month old were
carried out and the results are shown in Table 2.3. The open graded pavements have
a noise reduction of 2 to 3 ½ dB in relation to the dense graded pavement. SPB meas-urements
have also been carried out when the pavements were 52 month old in 2006
showing noise reductions quite similar to the results in month 16 [ 14].
Table 2.3. Noise measured by the Statistical Pass- By method for mixed traffic ( SPBI for passenger
cars at 96 km/ h ( 60 mph) and trucks at 88 km/ h ( 55 mph)) when the pavements were 16 month
old at a microphone height of 1.5 m and a distance of 7.5 m [ 11].
Pavement type SPBI
Month 16
Noise reduction relative
to DGAC month 16
DGAC 82.5 dB -
OGAC 75 79.0 dB 3.5 dB
OGAC 30 80.7 dB 1.7 dB
RAC- O 80.2 dB 2.3 dB
BWC 80.7 dB 1.8 dB
One of the authors has performed a visual inspection of the pavements in October
2008. The outcome of this inspection can be seen in Table 2.1. Generally the pave-ments
were found to be in a reasonable condition with no remarkable signs of wear
and tear that can have a significant influence of the noise emission except for cracks
in the BWC pavement.
38
2.2 Californian pavements
Another set of test pavements are included in this temperature project. These pave-ments
are part of a UCPRC project on noise emission from typical pavements used
in California [ 12] that is carried out for Caltrans as a part of the Caltrans Quieter Pave-ments
Research Work Plan. Two other pavements on roads in Davis are also included,
Old Davis Road ( ODR) and Road 105 ( RD 105).
Figure 2.2. Close up photos of eight of the Californian pavements included in this temperature
project. The diameter of the US quarter dollar coin is 24 mm.
DGAC QP- 9 DGAC QP- 43
OGAC QP- 28 OGAC QP- 4
OGAC QP- 3 RAC- G QP- 5
RAC- G QP- 31 RAC- G QP- 2
39
Three types of pavements were included:
• Dense Graded Asphalt Concrete ( DGAC). Four different pavements.
• Open Graded Asphalt Concrete ( OGAC). Three different pavements.
• Gap- graded rubberized asphalt concrete ( RAC- G). Three different pavements.
An overview of characteristics for the ten pavements can be seen in Table 2.4. Photos
of eight of the pavements can be seen in Figure 2.2. Further information is available
in [ 12].
Table 2.4. Data on the Californian pavements included in this temperature project [ 12].
Type Name Age in years Nominal Maxi-mum
Aggregate
Size
Air void
DGAC ODR - - -
DGAC QP- 9 7 12.5 mm 2.9 %
DGAC RD 105 - - -
DGAC QP- 43 2 12.5 mm 4.9 %
OGAC QP- 28 5 12.5 mm 12.8 %
OGAC QP- 4 5 12.5 mm 17.4 %
OGAC QP- 3 7 12.5 mm 19.2 %
RAC- G QP- 5 10 12.5 mm 8.0 %
RAC- G QP- 31 6 12.5 mm 7.3 %
RAC- G QP- 2 6 12.5 mm 9.3 %
40
41
3. The OBSI measurement method
The noise measurements have been performed using the On Board Sound Intensity
method ( OBSI) [ 1] as it is set up in the UCPRC Dodge Stratus sedan OBSI measure-ment
vehicle ( see Figure 3.1). The steel box behind the vehicle is an inertial laser pro-filometer
that measures the pavement elevation profile on both wheel tracks. The sur-face
texture expressed as the Mean Profile Depth ( MPD) is also measured in the right
wheel track. The OBSI measurement equipment has been developed by Paul Donavan
from the company Illingworth & Rodkin, Inc. in California.
Figure 3.1. The UCPRC OBSI measurement vehicle.
In the OBSI method, the sound intensity is measured. Sound intensity is a vector quan-tity
as it has both magnitude and direction. The sound intensity in a specified direction
is the amount of sound energy flowing through a unit area normal to that direction
[ 13]. It is a measure commonly used to measure the sound power of a given noise
source because the method can be used to focus on one noise source without interfer-ence
of noise from other sources.
Two sets ( probes) of two microphones are in the OBSI method placed at the leading
and the trailing edge of the right back tire ( passenger side). The microphones ( see Fig-ure
3.2) are placed 3 inches ( 76.2 mm) over the pavement surface and 4 inches ( 101.
6 mm) from the side of the tire. The distance between the two sets of microphones is
8.25 inches ( 209.6 mm). The sound intensity is measured in dB and the results are
A- weighted.
The OBSI measurements are performed at a speed of 60 mph ( 96 km/ h) on a pave-ment
section at a length of 134 m ( 5 seconds at 60 mph). The measurement is repeated
three times on the same pavement section.
42
The starting of a pavement section is marked on the road surface with reflective tape
or at the roadside by reflecting material mounted on a marking post. When a light ray
from the vehicle is reflected by the reflecting material a photo cell triggers the noise
measurement. The result is the average value of the three rounds of measurements on
the same pavement section.
Figure 3.2. The microphones for the intensity probe and the probe positions of the OBSI method.
In the CPX method currently used in Europe [ 2], the sound pressure level is measured
at fixed positions. The sound pressure level is measured in dB and the results are A-weighted
and averaged for the front and rear position. The position of the two micro-phones
in the CPX method are 100 mm ( 3.94 inches) over the pavement surface and
200 mm ( 7.87 inches) from the side of the tire. The distance between the two sets of
microphones is 400 mm ( 15.75 inches). The distance between the tires and the micro-phones
is twice as long in the CPX method as in the OBSI method. In the CPX
method, it is recommended to perform measurements of pavement sections with a
length of at least 100 m and at least 200 m in total shall be measured. The DRI- DK
application of the CPX method in an open trailer ( deciBellA) can be seen in [ 5].
Table 3.1. Microphone positions in the OBSI and CPX methods [ 2].
Method Distance to tire Height over pavement Distance between
microphones
OBSI 101.6 mm 76.2 mm 209.6 mm
CPX 200 mm 100 mm 400 mm
In different measurements, it has been found that OBSI levels normally are 2 to 4 dB
higher than CPX levels measured on the same pavement depending on which test tires
are used [ 17 and 22]. The higher OBSI levels can partly be explained by the micro-phone
positions where the OBSI microphones are placed much closer to the noise
source ( tire and pavement) than the CPX microphones. The different types of tires
used for CPX and the OBSI also explain the difference. It must be expected that the
two methods will rank pavements in the same way in relation to noise.
43
The following instruments and procedures were used for the temperature measure-ments.
A pocket weather station is used to measure air temperature. The measure-ments
are taken on the tested traffic lane at 1.2 to 1.5 meters over the pavement
( measurements out of the car’s window). A piece of paper/ cardboard was held over
the weather station in order to provide shielding from direct sun rays. The pavement
temperature is measured using a thermal infra- red gun, and is the average of three to
five readings taken on the right wheel path. Air and pavement temperature are meas-ured
immediately before and immediately after the OBSI testing. The devices for air
and pavement temperature are shown in Figure 3.3. The pocket weather station pro-vides,
in addition to air temperature, the air relative humidity and the barometric pres-sure.
Figure 3.3. Pocket weather station and Fluke thermal infrared gun used to measure respectively air
and pavement temperature.
In the expression for calculating sound intensity from sound pressures measured at
two closely spaced points, the actual air density is required. In most commercial ana-lyzers,
this is accounted for by entering the ambient air temperature and barometric
pressure at the time the data is acquired. If air density is not accounted for at the time
of the measurement, it can be accounted for afterwards by applying a correction factor
using the following formulas [ 23, 24]:
Mskg = 3.884266 × 10 ^ (( 7.5 × Tc)/( 237.7 + Tc))
Mkg = Mskg × Humidity%/ 100
Tvc = (( 1 + 1.609 × Mkg)/( 1 + Mkg)) × Tc
AirDensity = ( Baro × 100)/(( Tvc + 273) × 287))
OBSICorrection = 10 × ( Log10( ReferenceAirDensity) – Log10( AirDensity))
44
Mskg = factor to use in humidity correction
Tc = temperature (° C)
Mkg = adjustment for humidity
Baro = pressure in mbars
Tvc = application of correction to temperature using the humidity adjustment
ReferenceAirDensity = 1.21 kg/ m3
If not corrected by the noise analyzer used, the correction factor ( OBSICorrection) has
to be added to the measured sound intensity levels at each frequency. It can be seen in
the above formulas that the air temperature is included in the correction formulas. Fig-ure
3.4 shows the correction factor predicted for different temperatures and with all
other factors kept constant at the average levels measured when OBSI measurements
were performed at the LA138 test sections ( see Section 4) in February 2008 ( pressure
27.20 and relative humidity 40.5 %). It can be seen that the correction factor is 0 dB at
17 ° C ( 63° F) and that it increases with higher temperatures. At 30 ° C ( 86° F) it is
+ 0.29 dB.
- 1,000
- 0,500
0,000
0,500
1,000
30 50 70 90 110 130 150
Air temperature [° F]
OBSI Correction [ dB]
- 1,000
- 0,500
0,000
0,500
1,000
0 10 20 30 40 50 60
Air temperature [° C] OBSI Correction [ dB]
Figure 3.4. The correction factor ( OBSI Correction ) in dB for air density applied to OBSI sound intensity
measurements as a function of temperature ( in degrees Celsius to the right and Fahrenheit to the
left).
It will be noted that temperature is an issue for OBSI measurements of tire/ pavement
noise for two unrelated reasons.
1. Air density is a fundamental parameter in the determination of sound intensity.
If temperature and barometric pressure are not accounted for in the analyzer
used for the measurement, a correction can be applied if needed afterward.
Temperature is not a parameter for measurements of sound pressure ( like the
CPX and SPB/ CPB methods).
2. The temperature’s influence on results of noise measurements is related to the
mechanisms generating the noise when the tire rolls on the pavement. As dis-cussed
in Chapter 1, temperature can have an effect on the properties of the tire
and maybe also of the pavement.
45
4. The LA138 measurements
OBSI measurements were performed by UC Davis at the LA138 test sections ( see
Section 2.1) using the SRTT. The rubber hardness of the tire was measured to
67 Shore A some months before and after the time of the measurements at 24 ° C.
The noise measurements were carried out on February 26 to 28, 2008 in the daytime
between 5 am and 4 pm. Measurements were performed five times during the day at
each pavement in order to cover respectively low, medium, and high temperature. The
measurements were performed in both the westbound and eastbound direction on each
pavement except the DGAC, where the measurements for practical reasons were only
performed in one direction. The air and the pavement temperature were measured im-mediately
before and immediately after the noise testing. Each measurement was re-peated
three times one after another ( at practically the same temperature). The results
for one of the nine pavement sections are presented in the following – normalized by
application of the air density corrections ( see Chapter 3). The actual air density correc-tion
was between + 0.7 and + 1.1 dB.
Table 4.1. Normalized results of individual OBSI runs on the OGAC 75 pavement at different air
temperatures in the eastbound direction. The Standard Deviation for the three runs at the same
temperature is shown.
Temperature
in ° C
Run 1
OBSI in dB
Run 2
OBSI in dB
Run 3
OBSI in dB
Average
OBSI in dB
Standard
Deviation in dB
1.7 – 2.5 100.5 100.5 100.8 100.7 0.15
6.5 – 7.6 100.7 100.9 101.0 100.9 0.15
19.2 – 19.7 100.1 100.3 100.0 100.1 0.17
20.3 – 20.4 100.4 99.1 100.4 100.1 0.72
21.4 – 21.4 100.3 100.7 100.7 100.6 0.25
0,0
0,5
1,0
1,5
2,0
2,5
3,0
0,0 5,0 10,0 15,0 20,0 25,0
Air Temperature in degree- C
Standard Diviation in dB
Figure 4.1. Standard Deviation for all the sets of three repeated OBSI runs on the same pavement
at the same temperature.
46
As an illustration, the variations of the OBSI results for the three repeated measure-ments
carried out at the same pavement at the same temperature just after one another
are shown in Table 4.1 for the OGAC 75 pavement in the eastbound direction together
with the Standard Deviation.
In Figure 4.1, the Standard Deviation for all the sets of three OBSI runs on the same
pavement at the same temperature is presented for the temperature range of the meas-urements.
The Standard Deviation for the three repeated measurements at the same
temperature is generally below 0.5 dB with a few exceptions. There are two outliers
at 1.0 and 2.7 dB. These outlying data have not been included in the further analyses.
The Standard Deviation below 0.5 dB must be considered a reasonably good repeat-ability
of the OBSI measurements, but it is in the same order of magnitude as the tem-perature
influence on the noise ( see the following results). This highlights the general
problem of conducting noise measurements in order to investigate very small differ-ences
of noise levels. The current experiment contains a large series of OBSI meas-urements
which ensures reasonable reliability in the results.
One of the reasons for the variation in measured noise level at the three repeated runs
can be that the driver of the measurement car does not always hit exactly the same
wheel track or the same part of the wheel track with the right rear tire where the mi-crophones
are situated. Another reason can be minor uncertainty in the measurement
system used specially with regard to small speed variations as speed corrections were
not applied to the results.
The results are presented first in relation to the air temperature and afterwards in rela-tion
to the pavement temperature. The general relation between the air and pavement
temperature at these measurements can be seen in Figure1.2 and 1.3 in Section 1.1.
4.1 Air temperature and noise
The results of the measurements of noise and air temperature are shown in the follow-ing
figures for each of the pavements included in the project for both directions ( east-bound
and westbound). The air temperature was in the range from 2 to 22 ° C. The fig-ures
to the left show the normalized results of each of all the OBSI runs ( three per
pavement per temperature). A linear regression analysis is included. The figure to the
right shows the 1/ 3 octave band spectra at different temperatures as average spectra
for the three OBSI runs per pavement at approximately the same temperature.
47
y = - 0,032x + 100,903
R2 = 0,306
99
100
101
102
103
104
105
- 2 3 8 13 18 23
Air Temperature [ Degree- C]
OBSI [ dB]
70
75
80
85
90
95
100
500 800 1250 2000 3150 5000
Frequency [ Hz]
OBSI [ dB]
Air Temp= 2,1 C OBSI= 100,7 dB
Air Temp= 7,1 C OBSI= 100,9 dB
Air Temp= 19,4 C OBSI= 100,1 dB
Air Temp= 20,3 C OBSI= 100 dB
Air Temp= 21,4 C OBSI= 100,6 dB
Figure 4.2. OGAC 75 pavement eastbound. Normalized OBSI noise measurement results with
SRTT versus air temperature to the left and average spectra at the different temperatures to
the right.
The results from the OGAC 75 pavement in east and westbound directions can be seen
in Figure 4.2 and 4.3. The air temperature coefficients are - 0.032 dB/° C in both direc-tions.
Below 800 Hz, the frequency spectra are quite alike - independent of the tem-perature.
At the frequencies above 1000 Hz, the level is around 1 dB higher at 2 ° C
than at 20 ° C. The same tendencies were seen for different tires at the Sperenberg ex-periment
( see Figure 1.7 in Section 1.3).
y = - 0,032x + 102,041
R2 = 0,351
99
100
101
102
103
104
105
- 2 3 8 13 18 23
Air Temperature [ Degree- C]
OBSI [ dB]
70
75
80
85
90
95
100
500 800 1250 2000 3150 5000
Frequency [ Hz]
OBSI [ dB]
Air Temp= 2,4 C OBSI= 102,1 dB
Air Temp= 7,6 C OBSI= 101,7 dB
Air Temp= 19,5 C OBSI= 101,3 dB
Air Temp= 20,4 C OBSI= 101,5 dB
Air Temp= 21,4 C OBSI= 101,4 dB
Figure 4.3. OGAC 75 pavement westbound. Normalized OBSI noise measurement results with
SRTT versus air temperature to the left and average spectra at the different temperatures to
the right.
The noise levels in the westbound direction are around 1 dB higher than in the east-bound
direction on the same pavement. This general tendency is seen for all the four
pavements for which measurements have been carried out in both directions. In Table
4.2 it can be seen that the Medium Profile Depth ( MPD) is lower in the west direction
than in the east direction indication that the pavements are denser in the surface struc-ture
in the west direction and this can effect the noise generation. The asphalt pave-ments
were seven years old when the OBSI measurements were carried out. Differ-ences
in construction conditions and/ or tear and wear by traffic might be an explana-tion
for the difference on the same pavement between the east and westbound direc-tions.
This east/ west phenomenon has no influence on the temperature dependency
of the measurement results.
48
Table 4.2. Medium Profile Depth ( MPD) in Microns of the LA138 pavements in east-/ westbound
direction.
Direction OGAC 75 OGAC 30 RAC- O BWC DGAC
East 1054 997 815 726 -
West 967 887 686 714 745
y = - 0,032x + 100,837
R2 = 0,498
99
100
101
102
103
104
105
- 2 3 8 13 18 23
Air Temperature [ Degree- C]
OBSI [ dB]
70
75
80
85
90
95
100
500 800 1250 2000 3150 5000
Frequency [ Hz]
OBSI [ dB]
Air Temp= 2,1 C OBSI= 100,6 dB
Air Temp= 7,1 C OBSI= 100,8 dB
Air Temp= 19,4 C OBSI= 100,2 dB
Air Temp= 20,3 C OBSI= 100,2 dB
Air Temp= 21,4 C OBSI= 100,1 dB
Figure 4.4. OGAC 30 pavement eastbound. Normalized OBSI noise measurement results with
SRTT versus air temperature to the left and average spectra at the different temperatures to the
right.
y = - 0,030x + 102,410
R2 = 0,648
99
100
101
102
103
104
105
- 2 3 8 13 18 23
Air Temperature [ Degree- C]
OBSI [ dB]
70
75
80
85
90
95
100
500 800 1250 2000 3150 5000
Frequency [ Hz]
OBSI [ dB]
Air Temp= 2,4 C OBSI= 102,4 dB
Air Temp= 7,6 C OBSI= 102,1 dB
Air Temp= 19,5 C OBSI= 101,9 dB
Air Temp= 20,4 C OBSI= 101,9 dB
Air Temp= 21,4 C OBSI= 101,6 dB
Figure 4.5. OGAC 30 pavement westbound. Normalized OBSI noise measurement results with
SRTT versus air temperature to the left and average spectra at the different temperatures to
the right.
The results for the OGAC 30 pavement in the two directions are shown in Figure 4.4
and 4.5. The results of the temperature influence on the noise are quite similar to what
was seen for the OGAC 75 pavement. The air temperature coefficients for the two di-rections
are - 0.032 dB/° C and - 0.030 dB/° C respectively.
49
The results for the RAC- O pavement can be seen in Figure 4.6 and 4.7. The tempera-ture
coefficients are - 0.009 dB/° C and - 0.020 dB/° C respectively for the two directions
and less pronounced than for the two OGAC pavements.
y = - 0,009x + 100,841
R2 = 0,055
99
100
101
102
103
104
105
- 2 3 8 13 18 23
Air Temperature [ Degree- C]
OBSI [ dB]
70
75
80
85
90
95
100
500 800 1250 2000 3150 5000
Frequency [ Hz]
OBSI [ dB]
Air Temp= 2,9 C OBSI= 100,8 dB
Air Temp= 16,8 C OBSI= 100,9 dB
Air Temp= 20,7 C OBSI= 100,7 dB
Air Temp= 20,5 C OBSI= 100,5 dB
Air Temp= 21 C OBSI= 100,6 dB
Figure 4.6. RAC- O pavement eastbound. Normalized OBSI noise measurement results with SRTT
versus air temperature to the left and average spectra at the different temperatures to the right.
y = - 0,020x + 102,114
R2 = 0,377
99
100
101
102
103
104
105
- 2 3 8 13 18 23
Air Temperature [ Degree- C]
OBSI [ dB]
70
75
80
85
90
95
100
500 800 1250 2000 3150 5000
Frequency [ Hz]
OBSI [ dB]
Air Temp= 3 C OBSI= 102 dB
Air Temp= 17,3 C OBSI= 102 dB
Air Temp= 20,7 C OBSI= 101,6 dB
Air Temp= 21 C OBSI= 101,6 dB
Air Temp= 20,5 C OBSI= 101,7 dB
Figure 4.7. RAC- O pavement westbound. Normalized OBSI noise measurement results with SRTT
versus air temperature to the left and average spectra at the different temperatures to the right.
For the BWC pavement, the results can be seen in Figure 4.8 and 4.9. The temperature
coefficients for the two directions are - 0.013 dB/° C and - 0.029 dB/° C respectively.
This temperature coefficient is somewhat between the coefficients for the OGAC and
the RAC- O pavements.
50
y = - 0,013x + 103,493
R2 = 0,331
99
100
101
102
103
104
105
- 2 3 8 13 18 23
Air Temperature [ Degree- C]
OBSI [ dB]
70
75
80
85
90
95
100
500 800 1250 2000 3150 5000
Frequency [ Hz]
OBSI [ dB]
Air Temp= 2,9 C OBSI= 103,4 dB
Air Temp= 16,8 C OBSI= 103,5 dB
Air Temp= 20,7 C OBSI= 103,2 dB
Air Temp= 20,5 C OBSI= 103,1 dB
Air Temp= 21 C OBSI= 103,2 dB
Figure 4.8. BWC pavement eastbound. Normalized OBSI noise measurement results with SRTT ver-sus
air temperature to the left and average spectra at the different temperatures to the right.
y = - 0,029x + 104,037
R2 = 0,387
99
100
101
102
103
104
105
- 2 3 8 13 18 23
Air Temperature ([ Degree- C]
OBSI [ dB]
70
75
80
85
90
95
100
500 800 1250 2000 3150 5000
Frequency [ Hz]
OBSI [ dB]
Air Temp= 3 C OBSI= 103,9 dB
Air Temp= 17,3 C OBSI= 103,9 dB
Air Temp= 20,7 C OBSI= 103,4 dB
Air Temp= 20,5 C OBSI= 103,3 dB
Air Temp= 21 C OBSI= 103,3 dB
Figure 4.9. BWC pavement westbound. Normalized OBSI noise measurement results with SRTT ver-sus
air temperature to the left and average spectra at the different temperatures to the right.
The DGAC pavement was only measured in the westbound direction. The air tempera-ture
coefficient is - 0.046 dB/° C and higher than for the other pavements.
y = - 0,046x + 104,035
R2 = 0,558
99
100
101
102
103
104
105
- 2 3 8 13 18 23
Air Temperature [ Degree- C]
OBSI [ dB]
70
75
80
85
90
95
100
500 800 1250 2000 3150 5000
Frequency [ Hz]
OBSI [ dB]
Air Temp= 3 C OBSI= 103,9 dB
Air Temp= 17,2 C OBSI= 103,1 dB
Air Temp= 20,7 C OBSI= 103,2 dB
Figure 4.10. DGAC pavement westbound. Normalized OBSI noise measurement results with SRTT
versus air temperature to the left and average spectra at the different temperatures to the right.
51
- 0,050
- 0,045
- 0,040
- 0,035
- 0,030
- 0,025
- 0,020
- 0,015
- 0,010
- 0,005
0,000
OGAC 75
East
OGAC 75
West
OGAC 30
East
OGAC 30
West
RAC- O
East
RAC- O
West
BWC
East
BWC
West
DGAC
West
Temperature Coefficient c [ dB/ oC]
Figure 4.11. Air temperature coefficients in dB/° C measured in the range 2 – 22 ° C at LA138 using
the OBSI method and the SRTT.
The air temperature coefficients for all the nine measurements ranges between - 0.009
dB/° C and - 0.046 dB/° C ( see Figure 4.11). For the different pavement types the results
are the following:
• The average air temperature coefficient c is - 0.027 dB/° C ( or - 0.015 dB/° F) for all
nine measurements.
• For the dense pavements ( DGAC and BWC with air void respectively 7 and 5 %)
the average is - 0.029 dB/° C (- 0.016 dB/° F)
• For the open graded pavements ( OGAC 30, OGAC 75 and RAC- O all with an air
void of 10 to 11 %) the average is - 0.026 dB/° C (- 0.014 dB/° F).
All the data collected with the OBSI method on the LA138 test sections using an
SRTT in the air temperature range from 2 to 22 ° C.
4.2 Pavement temperature and noise
The following figures show the results of the OBSI noise measurements versus the
pavement temperature. The spectra are not shown since they are the same as shown
in Figures 4.2 to 4.10 just related to the pavement temperatures.
y = - 0,018x + 100,770
R2 = 0,250
99,0
100,0
101,0
102,0
103,0
104,0
105,0
- 2 3 8 13 18 23 28 33 38
Pavement temperature [ Degree- C]
OBSI [ dB]
y = - 0,018x + 101,908
R2 = 0,299
99,0
100,0
101,0
102,0
103,0
104,0
105,0
- 2 3 8 13 18 23 28 33 38
Pavement temperature [ Degree- C]
OBSI [ dB]
Figure 4.12. OGAC 75 pavement eastbound to the left and westbound to the right. Normalized
OBSI noise measurement results with SRTT versus pavement temperature.
52
y = - 0,021x + 100,761
R2 = 0,561
99,0
100,0
101,0
102,0
103,0
104,0
105,0
- 2 3 8 13 18 23 28 33 38
Pavement temperature [ Degree- C]
OBSI [ dB]
y = - 0,019x + 102,324
R2 = 0,695
99,0
100,0
101,0
102,0
103,0
104,0
105,0
- 2 3 8 13 18 23 28 33 38
Pavement temperature [ Degree- C]
OBSI [ dB]
Figure 4.13. OGAC 30 pavement eastbound to the left and westbound to the right. Normalized
OBSI noise measurement results with SRTT versus pavement temperature.
y = - 0,008x + 100,860
R2 = 0,154
99,0
100,0
101,0
102,0
103,0
104,0
105,0
- 2 3 8 13 18 23 28 33 38
Pavement temperature [ Degree- C]
OBSI [ dB]
y = - 0,013x + 102,057
R2 = 0,568
99,0
100,0
101,0
102,0
103,0
104,0
105,0
- 2 3 8 13 18 23 28 33 38
Pavement temperature [ Degree- C]
OBSI [ dB]
Figure 4.14. RAC- O pavement eastbound to the left and westbound to the right. Normalized OBSI
noise measurement results with SRTT versus pavement temperature.
y = - 0,010x + 103,477
R2 = 0,622
99,0
100,0
101,0
102,0
103,0
104,0
105,0
- 2 3 8 13 18 23 28 33 38
Pavement temperature [ Degree- C]
OBSI [ dB]
y = - 0,020x + 103,962
R2 = 0,619
99,0
100,0
101,0
102,0
103,0
104,0
105,0
- 2 3 8 13 18 23 28 33 38
Pavement temperature [ Degree- C]
OBSI [ dB]
Figure 4.15. BWC pavement eastbound to the left and westbound to the right. Normalized OBSI
noise measurement results with SRTT versus pavement temperature.
y = - 0,031x + 103,782
R2 = 0,448
99,0
100,0
101,0
102,0
103,0
104,0
105,0
- 2 3 8 13 18 23 28 33 38
Pavement temperature [ Degree- C]
OBSI [ dB]
Figure 4.16. DGAC pavement westbound. Normalized OBSI noise measurement results with SRTT
versus pavement temperature.
53
- 0,050
- 0,045
- 0,040
- 0,035
- 0,030
- 0,025
- 0,020
- 0,015
- 0,010
- 0,005
0,000
OGAC 75
East
OGAC 75
West
OGAC 30
East
OGAC 30
West
RAC- O East RAC- O West BWC East BWC West DGAC West
Temperature Coefficient c [ dB/ oC]
Figure 4.17. Pavement temperature coefficients in dB/° C measured in the range - 2 – 33 ° C at
LA138 using the OBSI method and the SRTT.
The pavement temperature coefficients for all the nine measurements range between
- 0.008 dB/° C and - 0.031 dB/° C ( see Figure 4.17). For the different pavement types
the results are the following:
• The average pavement temperature coefficient c for all nine pavements is -
0.018 dB/° C ( or - 0.010 dB/° F).
• For the dense pavements ( DGAC and BWC) the average is - 0.020 dB/° C
(- 0.011 dB/° F).
• For the open graded pavements ( OGAC 30, DGAC 75 and RAC- O) it is -
0.016 dB/° C (- 0.009 dB/° F).
Again, all data were collected with the OBSI method on the LA138 test sections using
an SRTT.
54
55
5. The California measurements
The results of the noise measurements performed on ten different Californian pave-ments
( see Section 2.2) at different temperatures can be seen in the following graphs.
The OBSI measurements were performed using a Goodyear Aquatred tire ( see Figure
5.1) which was the previous standard for OBSI before the SRTT was adopted.
The rubber hardness of the Aquatred tire was measured to 69 Shore A at the period
of the measurements and at 24 ° C.
The measurements were performed June- July 2007, and the purpose was to determine
at that time the feasibility of using pavement temperature correction. Due to practical
reasons, only the pavement temperature data are available for reporting. No spectral
data are available for this measurement series. In order to investigate the influence of
temperature, noise measurements were carried out on the same pavement on the same
day at 3 different temperatures ( targeted air temperatures 15, 25, and 35° C).
The pavement sections included in this study correspond to a small subset of the
total of asphalt pavement sections monitored by the UCPRC [ 12]. The QP number of
each section can be tracked down to the database of material properties and pavement
performance data [ 12]. Air density corrections have been performed on these results
( see Chapter 3).
Figure 5.1. A Goodyear Aquatred tire.
The results for the ten pavements grouped according to pavement types can be seen in
Figure 5.2 to 5.4.
56
y = - 0,066x + 103,750
R2 = 0,762
99,0
100,0
101,0
102,0
103,0
104,0
105,0
10 13 16 19 22 25 28 31 34
Pavement Temperature [ Degree- C]
OBSI [ dB]
OGAC QP- 3
y = - 0,073x + 104,936
R2 = 0,917
99,0
100,0
101,0
102,0
103,0
104,0
105,0
10 13 16 19 22 25 28 31 34
Pavement Temperature [ Degree- C]
OBSI [ dB]
OGAC QP- 28
y = - 0,156x + 104,814
R2 = 0,996
99,0
100,0
101,0
102,0
103,0
104,0
105,0
10 13 16 19 22 25 28 31 34
Pavement Temperature [ Degree- C]
OBSI [ dB]
OGAC QP- 4
Figure 5.2. Three OGAC pavements. Normalized OBSI noise measurement results with Aquatred
tire versus pavement temperature.
y = - 0,059x + 104,637
R2 = 0,746
99,0
100,0
101,0
102,0
103,0
104,0
105,0
10 13 16 19 22 25 28 31 34
Pavement Temperature [ Degree- C]
OBSI [ dB]
DGAC ODR
y = - 0,036x + 104,783
R2 = 0,936
99,0
100,0
101,0
102,0
103,0
104,0
105,0
10 13 16 19 22 25 28 31 34
Pavement Temperature [ Degree- C]
OBSI [ dB]
DGAC QP- 09
y = - 0,010x + 104,543
R2 = 0,115
99,0
100,0
101,0
102,0
103,0
104,0
105,0
10 13 16 19 22 25 28 31 34
Pavement Temperature [ Degree- C]
OBSI [ dB]
DGAC RD 105
y = - 0,073x + 104,936
R2 = 0,917
99,0
100,0
101,0
102,0
103,0
104,0
105,0
10 13 16 19 22 25 28 31 34
Pavement Temperature [ Degree- C]
OBSI [ dB]
DGAC QP- 43
Figure 5.3. Four DGAC pavements. Normalized OBSI noise measurement results with Aquatred tire
versus pavement temperature.
57
y = - 0,124x + 105,688
R2 = 0,903
99,0
100,0
101,0
102,0
103,0
104,0
105,0
10 13 16 19 22 25 28 31 34
Pavement Temperature [ Degree- C]
OBSI [ dB]
RAC- G QP- 02
y = - 0,038x + 104,077
R2 = 0,686
99,0
100,0
101,0
102,0
103,0
104,0
105,0
10 13 16 19 22 25 28 31 34
Pavement Temperature [ Degree- C]
OBSI [ dB]
RAC- G QP- 05
y = - 0,073x + 103,578
R2 = 0,961
99,0
100,0
101,0
102,0
103,0
104,0
105,0
10 13 16 19 22 25 28 31 34
Pavement Temperature [ Degree- C]
OBSI [ dB]
RAC- G QP- 31
Figure 5.4. Three RAC- G pavements. Normalized OBSI noise measurement results with Aquatred
tire versus pavement temperature.
- 0,180
- 0,160
- 0,140
- 0,120
- 0,100
- 0,080
- 0,060
- 0,040
- 0,020
0,000
DGAC
ODR
DGAC
QP- 9
DGAC
RD 105
DGAC
QP- 43
OGAC
QP- 28
OGAC
QP- 4
OGAC
QP- 3
RAC- G
QP- 5
RAC- G
QP- 31
RAC- G
QP- 2 Temperature Coefficient c [ dB/ oC]
Figure 5.5. Pavement temperature coefficients in dB/° C measured in the range 11 – 34 ° C at the
Californian pavements using the OBSI method and the Aquatred tire.
The summary of the pavement temperature coefficients from the measurements per-formed
on different Californian pavements can be seen in Figure 5.5. The average
coefficients for the different pavement types can be seen in Table 5.1 together with
average for all the ten pavements.
58
The two dense pavement types ( DGAC and RAC- DG) have nearly the same coeffi-cient
- 0.045 dB/° C and - 0.041 dB/° C with an average of - 0.043 dB/° C. For the open
graded pavement type ( OGAC), the coefficient is - 0.099 dB/° C, which is twice as
much. But the higher average for the open graded pavements is caused by the OGAC
QP- 4 with a very high coefficient. If this pavement is not considered, the difference
between the open and the dense pavements is reduced to - 0.043 dB/° C versus - 0.070
dB/° C.
All data are collected with the OBSI method on different Californian roads using an
Aquatred tire.
Table 5.1. Average pavement temperature coefficients for the three pavement types using the
OBSI method and the Aquatred tire.
Pavement type Pavement temperature
coefficient in Celsius
Pavement temperature
coefficient in Fahrenheit
DGAC - 0.045 dB/° C - 0.025 dB/° F
RAC- G - 0.041 dB/° C - 0.023 dB/° F
Average dense ( DGAC and RAC- G) - 0.043 dB/° C - 0.024 dB/° F
OGAC - 0.099 dB/° C - 0.055 dB/° F
Average all 10 pavements - 0.060 dB/° C - 0.033 dB/° F
59
6. Discussion and conclusion
It can be discussed if the temperature coefficient shall be applied in relation to the air,
pavement or tire temperature. There has so far been an international trend to use the
air temperature as the relevant variable, so this will be done in the general comparison
of results in the following.
One of the main objectives of this project was to investigate the influence of tempera-ture
on OBSI measurements as they are carried out with the methods currently used
in California. Here the SRTT is now used as the standard reference tire. Some years
ago the Aquatred tire was normally used. The temperature coefficients will also be
relevant for the CPX method when the SRTT is applied!
Table 6.1. Average temperature coefficients for the SRTT and the Aquatred tire on different as-phalt
pavement types ( LA138 and California measurements). The air temperature coefficient for
the Aquatred tire is predicted based on the ratio between air and pavement temperature coeffi-cients
for the SRTT.
Tire Dense asphalt
pavements
( DGAC)
Open graded
asphalt pavements
( OGAC and RAC- O)
Average all
asphalt
pavements
SRTT air temperature - 0.029 dB/° C - 0.026 dB/° C - 0.027 dB/° C
Aquatred air temperature ( predicted) - 0.062 dB/° C - 0.160 dB/° C - 0.090 dB/° C
SRTT pavement temperature - 0.020 dB/° C - 0.016 dB/° C - 0.018 dB/° C
Aquatred pavement temperature - 0.043 dB/° C - 0.099 dB/° C - 0.060 dB/° C
The measurement results from Sperenberg on a series of sixteen different passenger
car tires show a large variation in the air temperature coefficients ranging from - 0.035
to - 0.130 dB/° C for dense and porous asphalt pavements. The measurements at the
LA138 test sections using the SRTT and the measurements on the ten Californian sec-tions
with the Aquatred tire shows a significant difference for the temperature coeffi-cients
for these two tires ( see Table 6.1). There is only pavement temperature data
available for the California measurements. In order to estimate also the air temperature
coefficient for the Aquatred tire, the ratio between air and pavement temperature coef-ficients
for the SRTT is used in Table 6.1. The table then shows air as well as pave-ment
temperature coefficients for these two measurement series.
The average air temperature coefficient for the Aquatred is estimated at - 0.090 dB/° C
which is three times higher than for the SRTT. This means that the Aquatred tire
is much more sensitive to temperature than the SRTT. The rubber hardness was in this
investigation slightly lower for the SRTT than for the Aquatred tire ( 67 versus 69
Shore A).
60
This might partly explain the difference in temperature coefficients but other tire
properties like the chemical composition of the rubber, the tread pattern and the depth
differences etc. might play a role. The two tires in the French experiment had higher
rubber hardness of 76.3 and 79.5 shore A and the average air temperature coefficient
for these two tires is - 0.080 dB/° C for asphalt pavements ( see Table 6.2).
The average air temperature coefficient for the SRTT is - 0.027 dB/° C which is practi-cally
the same found by Donavan/ Lodico (- 0.026 dB/° C).
The results from different international measurement series presented in Chapter 1
indicate that also the pavement type has an influence on the temperature coefficients
( see Table 6.2 and Figure 6.1) even though there is some variation in the temperature
coefficients for the same pavement type. Most of the measurement series have been
performed in temperature spans of 20 to 30 ° C. Two measurement series has been per-formed
in more narrow temperature spans of just 10 ° C including the Arizona LAeq
measurements.
The trends and ranking between the different pavement types are not very clear.
The cement concrete pavements have the lowest temperature coefficient in all except
one measurement series. The exception is the LAeq Arizona results where the coeffi-cient
for cement concrete pavements is the highest coefficient reported on any pave-ment
in all the measurement series. Different measurement methods have been used,
narrow temperature range, and different sizes of vehicle/ tire populations have been
included and this is presumably a large part of the explanation for the differences.
From Table 6.2 it can also be seen that the SRTT ( see the LA138 OBSI measure-ments)
have significantly lower temperature coefficients than the other tires and tire
populations included in the comparison. This indicates that the SRTT is not very sen-sitive
to temperature variations.
Table 6.2. Air temperature coefficients for different pavement types in the 5 different measure-ment
series presented in Chapter 1 given as average values of all the tires used for passenger cars
and the SRTT measurements on LA138.
Name of
measurements
Measurement
method
Air
temperature
range
Dense
asphalt
pavements
( DGAC)
Open graded
asphalt
pavements
( OGAC)
Average all
asphalt
pavement
types
Cement
concrete
pavements
Sperenberg CPB 16 tires 0 – 35 ° C - 0.090 dB/° C - 0.061 dB/° C - 0.076 dB/° C - 0.043 dB/° C
France CPB 2 tires 0 – 30 ° C - 0.100 dB/° C - 0.060 dB/° C - 0.080 dB/° C - 0.030 dB/° C
Donavan/ Lodico OBSI 2 tires 30 – 40 ° C - 0.064 dB/° C - 0.057 dB/° C
LA138/ OBSI OBSI SRTT 2 – 22 ° C - 0.029 dB/° C - 0.026 dB/° C - 0.027 dB/° C
Arizona LAEQ 29 – 39 ° C - 0.064 dB/° C - 0.130 dB/° C
LA138/ SPB SPB 8 – 32 ° C - 0.022 dB/° C - 0.049 dB/° C - 0.036 dB/° C
61
Rough averages of all these different results presented in Table 6.2 can be seen in
Table 6.3. The very high temperature coefficient for cement concrete pavements
(- 0.130 dB/° C) reported in the Arizona LAeq measurements is not included.
Average values of temperature coefficients for different measurement series including
many different tires have relevance in relation to noise measurements where many dif-ferent
vehicles/ tires are included like SPB and LAeq measurements.
From Table 6.3 it can be seen that there is no big difference between temperature cor-rections
for dense (- 0.061 dB/° C) and open graded asphalt pavements (- 0.052 dB/° C).
According to these data, an average air temperature coefficient of - 0.057 dB/° C for all
types of asphalt pavements can be predicted. The correction coefficients for cement
concrete pavements is - 0.043 dB/° C and lower than for asphalt concrete pavements.
It can be seen that the difference in temperature coefficients for different pavement
types almost vanishes when a lot of different tires are included ( see Table 6.3).
Generally these results are quite close to the coefficient of - 0.05 dB/° C for passenger
cars commonly used in Denmark and the Netherlands. The results are in accordance
with the generic temperature correction method suggested by Sandberg in 2004 and
also reasonably close to the coefficient used in the EU tire noise directive of - 0.06
dB/° C up to 20 ° C. These factors are approximately double those for the SRTT
on asphalt tested in California.
Table 6.3. Rough average of all air temperature coefficients for the different pavement types from
the SRTT measurements presented in Table 6.1 and all the data from the different measurements in
Table 6. 2. The results for the concrete pavement in Arizona are not included.
Dense asphalt
pavements ( DGAC)
Open graded asphalt
pavements ( OGAC)
Average all asphalt
pavement types
Cement concrete
pavements
- 0.061 dB/° C - 0.052 dB/° C - 0.057 dB/° C - 0.043 dB/° C
62
- 0,140
- 0,120
- 0,100
- 0,080
- 0,060
- 0,040
- 0,020
0,000
Dense
asphalt
Temperature coeficient c [ dB/ oC]
Sperenberg
France
Donavan/ Lodico
LA138/ SRTT
Arizona
LA138/ SPB
Average of all
Open
asphalt
All
asphalt Concrete
Figure 6.1. Air temperature coefficients for different pavement types in the five different meas-urement
series presented in Chapter 1 given as average values of all the tires used for passenger
cars and the SRTT measurements on LA138. In the average for cement concrete the Arizona data
are not included.
For OBSI, CPX and CPB measurements only one or a few selected tires are used.
The results measured in this project and data collected from other sources clearly
show that the above mentioned average temperature corrections ( Table 6.3) are not
the most relevant for measurement methods using specific tires. Correction coeffi-cients
related to the specific tires used seem more appropriate. If measurements at dif-ferent
pavement types show great variation in the temperature coefficient for a specific
type of measurement tire, it can be relevant to determine pavement type specific tem-perature
correction coefficient. But as it can be seen from Table 6.1 the SRTT has
nearly the same temperature coefficients for different asphalt pavement types so it
does not seem relevant to use pavement specific correction factors for this tire when
used on asphalt concrete pavements.
There has not been any data available to evaluate the temperature coefficient for the
SRTT used on cement concrete pavements. Table 6.2 indicates that the temperature
coefficient for cement concrete pavements is lower than for asphalt pavements. But as
mentioned above, the temperature coefficients of the SRTT is not so sensitive to
pavement type. It could anyway be a recommendation to perform a survey like the
LA138 study with the SRTT on a series of different concrete pavements in a desert lo-cation
with a large temperature variation over the day.
Based on this project, the temperature correction coefficients presented in Table 6.4
for asphalt pavements is suggested for the SRTT used in the OBSI method and the
CPX method. The temperature influences the noise differently at different frequencies;
therefore it could be relevant to apply frequency dependent correction coefficients.
Table 6.5 and 6.6 as well as Figure 6.2 shows third octave band correction coefficients
for the SRTT in degrees Celsius and Fahrenheit.
63
Table 6.4. Suggestion for temperature coefficients for the SRTT used at asphalt pavements in the
OBSI method and other methods using SRTT.
In Celsius In Fahrenheit
Air temperature correction - 0.027 dB/° C - 0.015 dB/° F
Pavement temperature correction - 0.018 dB/° C - 0.010 dB/° F
Table 6.5. Suggestion for third octave band correction coefficients in dB per degree Celsius for the
SRTT used at asphalt pavements in the OBSI method.
Third Octave Band
( Hz)
Air Temperature
Correction ( dB/° C)
Pavement Temperature
Correction ( dB/° C)
500 - 0.040 - 0.023
630 - 0.054 - 0.034
800 0.003 - 0.001
1000 - 0.023 - 0.014
1250 - 0.033 - 0.020
1600 - 0.068 - 0.043
2000 - 0.054 - 0.034
2500 - 0.042 - 0.026
3150 - 0.075 - 0.047
4000 - 0.067 - 0.043
5000 - 0.109 - 0.068
- 0,12
- 0,1
- 0,08
- 0,06
- 0,04
- 0,02
0
0,02
500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Third octave band frequency [ Hz]
Air temperature coefficient [ dB/ oC]
Figure 6.2. Suggestion for third octave band air temperature correction coefficients in dB per de-gree
Celsius for the SRTT used at asphalt pavements in the OBSI method.
64
Table 6.6. Suggestion for third octave band correction coefficients in dB per degree Fahrenheit for
the SRTT used at asphalt pavements in the OBSI method.
Third Octave Band
( Hz)
Air Temperature
Correction ( dB/° F)
Pavement Temperature
Correction ( dB/° F)
500 - 0.022 - 0.013
630 - 0.030 - 0.019
800 0.002 - 0.001
1000 - 0.013 - 0.008
1250 - 0.018 - 0.011
1600 - 0.038 - 0.024
2000 - 0.030 - 0.019
2500 - 0.023 - 0.014
3150 - 0.042 - 0.026
4000 - 0.037 - 0.024
5000 - 0.061 - 0.038
There is still a need for more research in order to understand the basic mechanisms
related to temperature which is important for the generation of tire pavement noise.
The following general conclusions can be drawn regarding temperature corrections to
noise measurements:
• The air temperature has an important influence on the tire/ road noise measurements
results.
• The dependence of tire/ road noise on temperature can be approximated by a linear
relation.
• The temperature coefficients vary significantly for different tire types.
• The temperature coefficients are generally smaller for truck tires than for
passenger car tires.
• At low frequencies, the temperature coefficient is low. At frequencies above 1000
Hz the temperature coefficient is higher.
• The temperature coefficient is different for different pavement types.
• The temperature coefficient seems to be higher for dense asphalt concrete than for
open/ porous asphalt pavement.
• The temperature coefficient seems to be lower for cement concrete pavements than
for asphalt concrete pavements.
• The difference in temperature coefficients for different asphalt pavement types al-most
vanishes when many different tires are included.
• Temperature coefficients have to be determined specifically for each measurement
method taking into consideration the specific test tire( s) or the tire population in-cluded
in the measurements.
65
References
1. Further Development of the Sound Intensity Method of Measuring Tire Noise
Performance of In- Situ Pavements. Illingworth & Rodkin, Inc. Report prepared
for the California Department of Transportation January 4, 2006. See:
http:// www. i80. dot. ca. gov/ hq/ env/ noise/ pub/ 2_ Probe_ SI_ Report_ 04Jan06. pdf.
2. ISO/ CD 11819- 2: 2000. " Acoustics – Measurement of the influence of road
surfaces on traffic noise – Part 2: The close- proximity method." 2000.
3. ISO 11819- 1: 1997. " Acoustics – Measurement of the influence of road
surfaces on traffic noise – Part 1: Statistical Pass- by method." 1997.
4. " De methode Cwegdek 2002 voor wegverkeersgeluid", CROW- publication
2002 ( in Dutch).
5. Optimized thin layers – urban roads – the Kastrupvej experiment. Danish Road
Institute/ Road Directorate, Technical Note 66, 2008. See: www. roadinstitute. dk
6. Tyre/ Road Noise Reference Book. Ulf Sandberg and Jerzy A. Ejsmont. IN-FORMEX
2002. See: http:// www. informex. info/.
7. Further analysis of the Sperenberg data. Towards a better understanding of the
process influencing tyre/ road noise. Report M+ P. MVM. 99.3.1 revision 1. 30th
November 2001. M+ P Consulting Engineers, the Netherlands.
8. Acoustic performance. Low noise road pavements. Danish Road Institute/ Road
Directorate, Technical Note 44, 2006. See: www. roadinstitute. dk
9. Homepage of the SILVIA project: http:// www. trl. co. uk/ silvia/.
10. Road Stiffness Influence on Rolling Noise: Parametric Study Using a Rolling
Tire Model. JF. Hamet and P. Klein. INRETS, the French National Institute for
Transport and Safety Research. The report silvia- inrets- 008- wp2 published in
July 2003.
11. April 2007 Status Report: Caltrans Thin Lift Study: LA138 Tire/ Pavement
Noise Study. Judith Rochat. Volpe Center Acoustics Facility ( VCAF). April 30,
2007.
12. Investigation of Noise, Durability, Permeability and Friction Performance
Trends for Asphalt Surface Types First and Second year Results. Research re-port
UCPRC- RR- 2007- 03, University of California Pavement research Center,
Davis and Berkeley, CA, February 2008.
13. Sound Intensity. Brüel & Kjær booklet. Revision September 1993.
14. Volpe Center Updates on Tire/ Pavement Noise Studies. Judith L. Rochat. U. S.
DOT, Volpe Center Acoustics Facility. Presentation at TRB ADC40 Summer
Meeting San Luis Obispo, CA. July 2007. See:
http:// www. adc40. org/ summer2007/ 22% 20JRochat_ TRB% 20July% 202007. pdf
66
15. Effect of Test Parameters on OBSI Measurements. Paul R. Donavan, Illing-worht
& Rodkin, Inc and Dana M. Lodico, ICF Jones & Stokes. Presentation
at TRB ADC40 Summer Meeting Key West, Florida. July 2008. See:
http:// www. adc40. org/ summer2008/ LodicoTRB08. pdf
16. Volpe Center Research and Activities Update. Judith Rochat and Aaron Hast-ings.
U. S. DOT / RITA. Volpe Center Acoustics Facility. Presentation at TRB
ADC40 Summer Meeting Key West, Florida. July 2008. See:
http:// www. adc40. org/ summer2008/ RochatTRB08- P1. pdf
17. Comparative Measurements of Tire/ Pavement Noise in Europe and the United
States. Noise Intensity Testing in Europe ( NITE) Study. State of California
Department of Transportation. Paul R. Donavan, Illingworth & Rodkin, Inc.
July 26, 2006. A summary can be found at:
http:// www. dot. ca. gov/ hq/ env/ noise/ pub/ NITE_ Article_ for_ Caltrans. pdf
18. Semi- generic temperature corrections for tyre/ road noise. Ulf Sandberg.
Proceedings Inter. Noise conference 2004 in Prague.
19. Temperature effect on tyre– road noise. Fabienne Anfosso- Lédeé and Yves
Pichaud, LCPC. Applied Acoustics 68 ( 2007) 1– 16.
20. Directive 2001/ 43/ Ec of The European Parliament And of The Council of 27
June 2001 amending Council Directive 92/ 23/ EEC relating to tyres for motor
vehicles and their trailers and to their fitting.
21. Personal communication with Fabienne Anfosso- Lédeé LCPC, France February
2009.
22. Measuring Tire- Pavement Noise at the Source. Paul Donavan and Dana M.
Lodico. NCHRP Report 630. National Cooperative Highway Research Pro-gram.
Transportation Research Board. Washington 2009.
23. Excel air- density- correction spread sheet from Erwin Kohler.
24. Personal communication with Paul Donavan, Illingworth & Rodkin, Inc. April
2009.