Conference PaperPDF Available

Improvements in the CPX method and its ability to predict traffic noise emissions

Authors:
  • Grolimund + Partner AG

Abstract and Figures

Recent studies have provided for substantial improvements to measurements made by the Close-Proximity (CPX) method, intended for measurement of noise properties of road surfaces. These new findings were incorporated in the new standards published in 2017: the ISO 11819-2 (the CPX method) and Technical Specifications 11819-3 (about the reference tyres) as well as 13471-1 (about temperature corrections). This study, firstly, investigates the typical uncertainties associated with speed, ambient temperature and rubber hardness corrections. Secondly, the study evaluates to what degree the new standards improved the measurement method's repeatability by evaluating measurements that were undertaken on the same road surface at different times. Thirdly, the paper assesses the method's ability to predict the effect of road surfaces on roadside traffic noise by analysing the relationship with statistical pass-by measurements (SPB) undertaken on a large number of road surfaces within the same time frame. The study shows that the repeatability of CPX-measurements could be significantly improved by the new ISO standards, while some uncertainties associated with the properties of the test tyres remain. The study, moreover, provides evidence that overall and spectral road side traffic noise emissions can be reliably predicted by the CPX-method.
Content may be subject to copyright.
Improvements in the CPX method and its ability to predict
traffic noise emissions
Bühlmann, Erik
1
Grolimund + Partner AG environmental engineering
Thunstrasse 101A, CH-3006, Bern
ABSTRACT
Recent studies have provided for substantial improvements to measurements made
by the Close-Proximity (CPX) method, intended for measurement of noise
properties of road surfaces. These new findings were incorporated in the new
standards published in 2017: the ISO 11819-2 (the CPX method) and Technical
Specifications 11819-3 (about the reference tyres) as well as 13471-1 (about
temperature corrections). This study, firstly, investigates the typical uncertainties
associated with speed, ambient temperature and rubber hardness corrections.
Secondly, the study evaluates to what degree the new standards improved the
measurement method’s repeatability by evaluating measurements that were
undertaken on the same road surface at different times. Thirdly, the paper assesses
the method’s ability to predict the effect of road surfaces on roadside traffic noise
by analysing the relationship with statistical pass-by measurements (SPB)
undertaken on a large number of road surfaces within the same time frame. The
study shows that the repeatability of CPX-measurements could be significantly
improved by the new ISO standards, while some uncertainties associated with the
properties of the test tyres remain. The study, moreover, provides evidence that
overall and spectral road side traffic noise emissions can be reliably predicted by
the CPX-method.
Keywords: Tyre/road noise, Measurement methods, Road surfaces
I-INCE Classification of Subject Number: 13, 72
1 INTRODUCTION
For speeds of 30 km/h and above the road surface is regarded as the most
important influencing factor regarding the generation of road traffic noise [1].
Consequently, the acoustic quality of a road surface has become an important input
factor in noise emission modelling. Low-noise road surfaces have become one of the
preferred measures to reduce excessive traffic noise in many countries, because of their
substantial noise reduction potential and area-wide effect, see e.g. [2, 3]. In many cases,
the acoustic performance of a road surface is subjected to success monitoring or
acoustic conformity testing (e.g. in cases where the noise reduction by a road surface is
specified in construction tenders and contracts). In such cases, the Close Proximity
1
erik.buehlmann@grolimund-partner.ch
(CPX) method is often used to test the conformity with the demanded values. The
advantages of the CPX method are that it is both cost-effective and capable of
evaluating the acoustic pavement characteristics along entire road sections. This gives
the CPX method a real advantage over the Statistical Pass-By method (SPB) [4] which
evaluates the acoustic properties of road surfaces in a cross-section by measurement of
statistical noise emissions of vehicle pass-bys. Recent pre-normative studies have
provided for substantial improvements of the CPX method. These new findings were
incorporated in the new standards published in 2017: the ISO 11819-2 [5] (about the
CPX method), the ISO/TS 11819-3 [6] (about the Reference Tyres) as well as the
ISO/TS 13471-1 [7] (about Temperature Corrections). Because of the CPX method’s
high degree of standardisation (i.e. involving controlled measurements using a well-
defined system and sets of test tyres) the method is thought to bring advantages
regarding measurement repeatability. The fact that CPX measurements cannot be
undertaken in entirely controlled environments, however, introduces a certain degree of
measurement uncertainty. Moreover, there remain questions regarding the degree that
the CPX test tyres are capable of predicting the noise emissions of a statistical tyre and
vehicle fleet on a specific road surface.
The objectives of this study are, firstly, to review the CPX method’s main
influencing factors and to estimate the typical uncertainties associated with the
correction schemes specified in the ISO standards; secondly, to assess the repeatability
of the method; thirdly, to examine the method’s ability to predict roadside traffic noise
emissions on a particular road surface. The study concludes by making suggestions for
further improvements of the method’s repeatability.
2 MATERIALS & METHODS
2.1 The measurement principle
The CPX method is characterised by driving over a road surface at constant
speed while continuously measuring the tyre/road noise emitted by standardised test
tyres at two defined microphone positions 20 cm from the tyre. This can be done either
with a one-wheeled or two-wheeled trailer (with isolated chamber(s) to shield the
measurement from background noise e.g. from free flowing traffic) or at the tyre of the
test vehicle itself. Reproducibility of the measurement is ensured by correcting the
measured noise levels for the influence of the measurement system by a spectral free-
field correction [see 5]. By applying a set of provided generic and semi-generic
correction factors for speed, ambient temperature and tyre rubber hardness, the method
ensures that the measured tyre/road noise values are corrected for the specific conditions
during a measurement [see 5, 6]. As the magnitudes of these influences may vary for
each tyre-pavement combination, every generic or semi-generic correction is subject to
a certain uncertainty.
2.2 Reviewing the main influencing factors on CPX-measurement results and
estimating the uncertainties associated with the correction schemes
Existing research on the topic suggests that the main influencing factors on CPX
measurement results are driving speed and ambient air temperature during the
measurements as well as tyre rubber hardness of the test tyres at the time of
measurement [e.g. 8, 9, 10]. These influencing factors are evaluated and further
discussed based on existing pre-normative research. The focus is, thereby, made on
research that specifically relates to the test tyres P1 (for passenger cars) and H1 (for
heavy vehicles) specified in ISO/TS 11819-3.
The corrections procedures required by the standards ISO 11819-2 and ISO
11819-3 aim at ensuring a high repeatability of the CPX-method. There, however,
remains a degree of uncertainty for each of the correction procedures. This is especially
the case when test conditions deviate considerably from the reference conditions
defined in the standards. In a first step, the analysis investigates the degree of variation
of these influencing factors and relates them to the correction factors provided in the
standards. Secondly, the CPX measurement database of Grolimund + partner AG
environmental engineering (G+P), comprising a total of 8’382 km of measurement runs
undertaken in the scope of monitoring measurements, is evaluated for typical variability
in the test conditions. The database stores the recorded signals during each measurement
run undertaken between 2008 and 2018 per road segment of 20 m. Thirdly, an estimate
of the typical measurement uncertainties related to the correction schemes is given.
2.3 Repeatability tests
The repeatability of CPX measurements is assessed by comparing the
measurement results on a range of test tracks undertaken at two different times within
the same year. The underlying assumption is that the test track’s surface properties did
not change during this period. In order to investigate the accuracy gains achieved by the
ISO standards published in 2017, each of these measurements is evaluated, firstly, by
applying the correction schemes of the 3rd Committee Draft (CD) of 2000 and,
secondly, by using the correction schemes of the current ISO standards.
2.4 Correlation between CPX and SPB measurements
To evaluate the ability of CPX test tyres to predict the effect of a road surface
for a statistical vehicle fleet, correlations between CPX and SPB measurements are
established. The analysis investigates to what degree SPB maximum overall and
spectral noise levels for passenger cars can be predicted based on CPX measurements
with the tyre P1 by undertaking regression analyses.
3 RESULTS & DISCUSSION
3.1 Main influencing factors and estimated uncertainties of correction schemes
3.1.1 Driving speed
One of the most important influencing factors of CPX measurement results is the
driving speed during the measurements. The ISO standard [5], therefore, requires
measurements to be undertaken at a certain reference speed (in the case of Switzerland
these are 50 on urban roads and 80 km/h on national roads). In practice, the exact
reference speed cannot always be guaranteed. For such cases, the standard allows a
correction for speeds that deviate by a maximum of ±15% from the reference speed on
each road segment and by a maximum of ±5% over a tested road section. Schwanen et
al. 2007 [8] investigated the speed coefficient of the CPX test tyre P1 on various road
surfaces. An overview of the speed coefficients obtained in [8] is given in Figure 1.
Figure 1 Speed coefficients on different road surfaces obtained by Schwanen et al.
2007 for porous road surfaces (blue) and dense surfaces (green) in comparison with the
speed coefficients B defined in the standard ISO 11819-2.
Figure 1 shows there is some degree of variation of the speed coefficient for
different road surfaces within the same road surface category. ISO 11819-2 provides
semi-generic speed coefficients B that shall be applied depending upon the road surface
category (r), i.e. B=25 for porous asphalts, B=30 for semi-porous or dense asphalts as
well as clogged porous asphalts, and B=35 for cement concrete surfaces. The speed
correction is applied as follows:
Cvref, r= Equation 1
where v is the actual driving speed and vref is the reference speed, B is the speed
coefficient and Cvref, r the correction for speed deviations from the reference speed vref.
The true influence of speed on tyre/road noise is dependent upon the underlying noise
generation mechanisms, each of which show a different speed dependency [11]. The
predominance of certain noise generation mechanisms in turn is dependent upon the
specific tyre-pavement combination. As a consequence, the speed coefficient can vary
by as much as ±5 units from the speed coefficient B specified in [5] for a particular road
surface within the same category (see Figure 1). In the worst case the speed correction
introduces an uncertainty of around 0.5 dB where measurements are undertaken with
the maximum allowed speed tolerance on a particular road segment (say at 42.5 km/h
instead of 50 km/h). In view of this, measurements whenever possible are undertaken as
close to the reference speed as possible to avoid unnecessary errors in the measurement
results. In order to estimate the typical uncertainty associated with the correction for
speed, the database of G+P for CPX measurements is evaluated regarding the
distribution of driving speed during the measurements undertaken at reference speeds of
50 and 80 km/h (see Figure 2).
25
30
35
15
20
25
30
35
40
45
ISO 11819-2 for porous s.
Average
PAC 25 mm + PAC 25 mm
PAC 25 mm + PAC 45 mm
PAC 25 mm + PAC 35 mm
PAC 25 mm + PAC 25 mm
PAC 0/11
PAC 45 mm + PAC 25 mm
PAC 50 mm
PAC 25 mm
PAC 0/16
PAC 25 mm
PAC 25 mm + PAC 45 mm
PAC 45 mm
PAC 25 mm
PAC 25 mm + PAC 45 mm
Thin Layered Asphalt 12%
PAC 45 mm + PAC 45 mm
Thin Layered Asphalt 12%
PAC 35 mm + PAC 55 mm
Thin Layered Asphalt 8%
PAC 25 mm + PAC 65 mm
PAC 25 mm + EPAC 45 mm
PAC 25 mm + PAC 25 mm
Thin Layered Asphalt 12%
PAC 25 mm + PAC 45 mm
PAC 25 mm + EPAC 45 mm
ISO 11819-2 for dense s.
Average
ISO 10844
SMA 0/11
SMA 0/16
SMA 0/8
SMA 0/6
DAC 0/16
surface dressing
surface dressing
ISO 11819-2 for cement concr.
speed coefficient B
road surface
Figure 2 Distribution of driving speed during 8’382 km of monitoring measurement
undertaken by G+P at reference speed 50 km/h (left) and 80 km/h (right) together with
mean μ and standard deviation ϭ.
As Figure 2 shows, the mean speeds during measurements undertaken by G+P
are virtually at the reference speed. The standard deviations derived for measurements
on urban roads (1.3 km/h) and for measurements on national roads (1.2 km/h) can be
considered as reasonably small. When combined with the variation of the speed
coefficients within the same road surface category, this would lead to a typical
uncertainty associated with the speed correction of around 0.05 dB(A).
3.1.2 Ambient temperature
Another important influencing factor on CPX measurement results is the
ambient air temperature during the measurement. Many authors found considerable
temperature effects with noise levels decreasing up to 1 dB(A) per 10 °C rise in air
temperature on dense asphalts, highlighting the need for temperature correction of
measurement results [e.g. 10, 12, 13, 14, 15]. Reliable temperature correction of
measured sound levels is therefore crucial. The working group ISO/TC43/SC1/WG27
compiled the temperature effects for the CPX test tyres P1 and H1 found in various
studies. The resulting data compilation is shown in Figure 3.
Figure 3 Data compilation of noise-temperature slopes by ISO/TC43/SC1/WG27 for
CPX test tyres P1 (left) and H1 (right) per road surface category
-0.08
-0.09
-0.07
-0.04
-0.16
-0.14
-0.12
-0.10
-0.08
-0.06
-0.04
-0.02
0.00
global
(n=64)
dense
(n=45)
cement
concrete
(n=5)
porous
(n=14)
noise-temperature slopes [dB(A)/°C]
tyre P1
-0.10 -0.10
-0.06 -0.05
-0.16
-0.14
-0.12
-0.10
-0.08
-0.06
-0.04
-0.02
0.00
global
(n=47)
dense
(n=36)
cement
concrete
(n=4)
porous
(n=7)
noise-temperature slopes [dB(A)/°C]
tyre H1
As illustrated in Figure 3, the temperature effects vary between different road
surface categories, while only marginal differences between the two CPX test tyres P1
and H1 were found. Similar to the speed behaviour investigated in section 3.1.1, noise-
temperature slopes vary for different road surfaces within the same category by as much
as ±0.02 dB(A)/°C (distance of quartiles from median temperature effect). The reason
for this variation is again, that the contributions of different noise generation
mechanisms can vary depending on the specific tyre/road surface combination. These
noise generation mechanisms in turn are not equally influenced by temperature. Within
the temperature range allowed in [5] this would lead to a maximum error of 0.3 dB(A).
Moreover, temperature effects vary also with speed [16]. The standard ISO/TS 13471-1
[7], therefore, provides speed dependent factors for each road surface category to
correct for the influence of ambient air temperature during the measurement. The
correction is made to a reference ambient air temperature of 20°C and is specified as
follows:
Equation 2
where γt is the temperature coefficient for tyre t (either P1 or H1), in dB/°C; T is
the air temperature during the CPX measurement, in °C; Tref is the reference air
temperature = 20.0 °C; CT,t is the CPX level correction for temperature (T) for tyre t, in
dB. While for dense asphaltic surfaces (such as DAC, SMA, TAL with air voids
typically below 18 %, and surface dressings) 𝛾P1=𝛾H1=−0.14+0.0006 𝑣, for cement
concrete surfaces of all types 𝛾P1=𝛾H1=−0.10+0.0004𝑣 and for porous asphalts a factor
of 𝛾P1=𝛾H1=−0.08+0.000 4𝑣 is applied. Often it is not possible to carry out CPX
measurements at the reference ambient air temperature of 20 °C. The database on CPX
measurements of G+P was analysed regarding the distribution of ambient air
temperatures occurring during the measurements. The results of this analysis are shown
in Figure 4.
Figure 4 Distribution of ambient air temperatures (T) in °C during 8’382 km of
monitoring measurements undertaken by G+P together with mean μ and standard
deviation ϭ.
According to Figure 4, the ambient air temperatures used for correction of CPX
measurements are nearly centred on the reference temperature of 20 °C. The majority of
measurements are undertaken at temperatures between 10 and 27 °C (i.e. within a range
of 10 °C from the reference temperature), a limited number of measurements are also
undertaken near the minimum and maximum temperature allowed by the ISO standard.
When combining the typical temperature range with the variation of the noise-
temperature slope within the same road surface category, this results in a typical
uncertainty of 0.2 dB(A) for temperature correction.
3.1.3 Rubber hardness of test tyres
A third important influencing factor on CPX measurement results is the rubber
hardness of the test tyres. Practical experience shows that test tyres can get significantly
harder within a single measurement season [17]. This is why ISO/TS 11819-3 [6]
includes a correction for tyre rubber hardness which was recently updated based on the
results of [18]. The study provided a compilation of tyre-rubber hardness influence on
CPX measurements based on various studies, which is presented in Figure 5.
Figure 5 Rubber hardness effect based on 12 datasets and 171 different relationships
for P1 tyre and H1 tyre and the rubber hardness correction specified in the updated
standard ISO 11819-3 (triangles), source [18].
The data compilation in Figure 5 shows a variability of the rubber hardness
effect within the same tyre of ±0.05 for tyre P1 and of up to 0.12 dB/Shore A for tyre
H1 (distance of quartiles from the standard correction factor). Within the rubber
hardness range for CPX test tyres allowed in [6] this would lead to a maximum error of
0.35 dB for tyre P1 tyre and of 0.84 dB for tyre H1. The procedure for correcting for
tyre-rubber hardness is specified as follows.
Equation 3
where βt is the rubber hardness coefficient for tyre t (β = 0.12 for the P1 tyre and
β = 0.20 for the H1 tyre in the updated version of ISO 11819-3), in dB/Shore A; HA is
the measured rubber hardness, in Shore A; Href is the reference rubber hardness = 66
Shore A; CHA,t is the CPX level correction for rubber hardness (HA) for tyre t, in dB.
As the test tyres’ rubber hardness increases at a rather fast rate during the
measurement season even when stored at cool and controlled conditions, it is often not
possible to carry out CPX measurements at the reference rubber hardness. The data base
0.13 0.12
0.17
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
global
(n=171)
SRTT
(n=96)
Avon
AV4
(n=75)
rubber harndess effect [dB/ Shore A]
tyre
P1
tyre
H1
on CPX measurements of G+P is analysed regarding the distribution of rubber hardness
of the P1 and H1 test tyres at the time the measurements. The dataset contains rubber
hardness values for all measurements undertaken between 2013 (the year since tyre
rubber hardness was systematically measured) and 2018. The results of this analysis are
shown in Figure 6.
Figure 6 Distribution of rubber hardness in Shore A of test tyres P1 (for passenger
cars) and H1 (for heavy vehicles) during all monitoring measurement undertaken by
G+P since 2013 together with mean μ and standard deviation ϭ. The dotted lines
represent the tolerances allowed in the ISO standard.
Figure 6 shows that rubber hardness of the test tyres varied typically between 65
and 70 Shore A for tyre P1 and between 68 and 72 Shore A for tyre H1. The rubber
hardness of both test tyres is not normally distributed around the reference rubber
hardness. If this rubber hardness range is combined with the typical variation of the
rubber hardness effect for the test tyre this results in a typical uncertainty for the rubber
hardness correction of 0.2 dB for tyre P1 and of 0.7 dB for tyre H1.
3.2 Repeatability of CPX measurements
In this section the repeatability of CPX method is assessed by evaluating the
difference of noise levels obtained from several separate measurement campaigns
undertaken at different times within the same year (in most cases with deviating
measurement conditions regarding speed, ambient air temperature and tyre rubber
hardness). Data from 9 different road sections and from two different years (with
different sets of test tyres) were incorporated in the analysis. The results from this
practical repeatability test are shown in Figure 7, together with the measurement
conditions (driving speed v, ambient air temperature T, rubber hardness S) experienced
during each of the measurements, while indicating the month the measurements were
undertaken. Please note that the underlying assumption is that the test track’s surface
properties did not change during the period between the measurement pairs. As all
measurements displayed in Figure 7 were undertaken on low-noise road surfaces, this
evaluation can be considered something like a worst case scenario, as the acoustic
properties of low-noise road surface often vary over the cross-section of the wheel track
(reducing repeatability if measurement runs do not take place in the exact same lateral
position). Moreover, this type of surfaces can experience faster changes in the acoustic
properties than other types of surfaces.
Figure 7 Difference in LCPX, P (left chart) and LCPX, H (right chart) for measurements
undertaken at two different times within the same year (see labels for measurement
conditions). Measurements were evaluated according to the correction schemes of the
current ISO standards (green) and the correction schemes of 3rd CD of 2000 (red).
As Figure 7 shows, the repeatability of CPX measurements was significantly
improved by the correction schemes of the ISO standards published in 2017. This first
evaluation yielded a repeatability of 0.45±0.3 dB for tyre P1 and of 0.35±0.3 dB for tyre
H1 under the current ISO standards. Hence, the repeatability could be improved by
0.5 dB (down from 0.95 dB) for tyre P1 and by 0.10 dB (down from 0.45 dB) for tyre
H1 if compared with the correction schemes of the 3rd CD dating back to the year 2000.
3.3 Remaining uncertainties
3.3.1 Lack of spectral correction approaches
It is known that the temperature effects and the influence of driving speed vary
over the noise spectra [e.g. 8, 7]. To a lesser degree this is also the case for tyre rubber
hardness influence [18]. Correction should, therefore, ideally be made based on spectra.
Currently available data, however, are not sufficiently consistent to introduce
frequency-dependent correction schemes in standardisation at this moment. For this
reason, the current ISO standards prescribe that the same correction factors are applied
for all frequencies. As more data on the spectral influences becomes available, a
spectral correction may further reduce uncertainty of the CPX method, ultimately
leading to improved repeatability of the method.
3.3.2 Acoustic conformity of test tyres
In study [18] the acoustic conformity of CPX test tyres with the same rubber
hardness was tested on a drum in the laboratory. The study revealed that the standard
deviations of noise levels for the P1 tyre are rather small and homogeneous over the
noise spectrum and lay near the expected measurement uncertainty. Tyre P1 is designed
to be used for various tests in the automotive industry. To ensure comparability of these
tests, the tyre is made available in the long-term to car manufacturers with nearly
unchanged rubber compounds. This leads also to a satisfactory acoustic conformity of
the tyre. The study indicated, however, that for the H1 tyre the variation of spectral
noise levels was considerably larger in some of the third-octave bands. This variability
can only partially be mitigated when tyres of the same production batch are used. In
contrast to tyre P1, the H1 tyre is a market tyre. According to information from tyre
manufacturers, it is a normal procedure for tyre manufacturers to adapt the rubber
compounds during the lifespan of a product line. In order to improve the repeatability of
measurements with the H1 tyre, one should conduct conformity tests (ideally with
several tyres and then select the most conformant one) before replacing a reference tyre.
Alternatively, a procedure may be developed to correct for this. More research is
planned on this topic with the aim of addressing the conformity issues of the H1 tyre in
the near future.
3.4 Reliability in predicting road-side noise levels
To assess the ability of CPX test tyre P1 to predict the effect of a road surface
for a statistical vehicle fleet of passenger cars, regression analysis for a total of 110 pairs
of CPX and SPB measurements derived on a total of different road surfaces and road
sections at 50 km/h was undertaken. The obtained correlations between LCPX,P and LSPB,P
for overall noise levels (LAFmax) is shown together with the 95% confidence intervals in
Figure 8. It needs to be highlighted that the measurements included in the sample were
not undertaken with the aim of establishing a correlation model and that all
measurements were selected for display (i.e. without checking deviations for free field
conditions required in [4]; these requirements are often difficult to meet for test sites in
urban areas).
Figure 8 Correlation between overall noise levels obtained for the CPX tyre P1 and
those obtained from SPB measurements for passenger cars at 50 km/h.
Figure 8 shows that with a correlation coefficient of 0.87 overall noise levels
between the two methods CPX and SPB correlate relatively well. The regression
analysis was repeated for spectral noise levels shown in Figure 9.
LCPX, P [dB(A)]
LSPB, P [dB(A)]
Figure 9 Correlation between spectral noise levels obtained from CPX measurements
for tyre P1 (x-axis) and those obtained from SPB measurements for passenger cars
(y-axis) displayed together with the 95% confidence intervals.
As Figure 9 shows, the spectral noise levels between CPX and SPB
measurements correlate relatively well for most third-octave bands with correlation
coefficients of 0.80 and higher (green plots). This is not equally true for third-octave
bands of 325 Hz, 500 Hz and 630 Hz (yellow and red plots). The reason for this may be
that these third-octave bands coincide with the block pattern frequencies specific to the
P1 tyre. Tyre related noise generation may vary between different tyres and, hence, may
not be identical for tyres of the statistical vehicle fleet. From the regression analysis
provided in Figure 8 and Figure 9 it can be concluded that CPX measurements are
suitable for predicting the road side effect of road surfaces even when spectral noise
levels are to be considered, for instance as input data in modern noise emission models
where spectral corrections for the properties of the road surface are required.
4 CONCLUSIONS
This study investigated the typical uncertainties associated with the correction
schemes of the CPX method, its repeatability as well as the methods ability to predict
the effect of road surfaces on roadside traffic noise. When applying the correction
schemes of the current ISO standards, it was found that tyre variation in the rubber
hardness correction contributed most (up to 0.7 dB) to the uncertainty of measurement
results. While the typical uncertainty associated with the correction of temperature
influence (up to 0.2 dB) is less significant and the typical uncertainty for the correction
of the speed influence (up to 0.05 dB) is close to negligible, if larger deviations from the
reference speed are avoided. Based on the limited practical repeatability test provided in
this study, a repeatability 0.45±0.3 dB was determined for the CPX method. The study,
moreover, provided evidence that overall (R2 = 0.87) and spectral (R2 > 0.80 for most
third-octave bands) road side traffic noise emissions can be reliably predicted by the
CPX method. In order to further increase the repeatability and reproducibility of the
CPX method, the author advises refining the rubber hardness correction for test tyres
315 Hz 500 Hz 630 Hz 800 Hz
2000 Hz
5000 Hz
1600 Hz
4000 Hz
1250 Hz
3150 Hz
1000 Hz
2500 Hz
(e.g. in establishing separate correction factors for different road surface categories) and
to establish a procedure to address the acoustic conformity issues of the H1 tyre.
5 ACKNOWLEDGEMENTS
The author is grateful to Felix Schlatter for performing data analysis and for
preparing and designing the illustrations of this paper.
6 REFERENCES
1 E. hlmann, S. Egger, E. Hammer and T. Ziegler, Basic information for assessing noise
effects at speed limit 30 km/h (in German), VSS2012/2014, Bern, Switzerland (2017).
2 H. Bendtsen, J. Kragh and E. Nielsen, „Use of noise reducing pavements - European
experience, Hedehusenen, Denmark (2008).
3 S. Steiner, E. Bühlmann and E. Hammer, „Milestones in establishing low-noise asphalts as a
widely used, effective and durable noise abatement measure in Switzerland”, Proc.
EURONOISE, Crete, Greece (2018).
4 ISO 11819-1, “Acoustics — Measurement of the influence of road surfaces on traffic noise
Part 1: The Statistical Pass-By method”, ISO, Geneva, Switzerland (1997).
5 ISO 11819-2, “Acoustics — Measurement of the influence of road surfaces on traffic noise
Part 2: The close-proximity method”, ISO, Geneva, Switzerland (2017).
6 ISO/TS 11819-3, “Acoustics — Measurement of the influence of road surfaces on traffic
noise Part 3: Reference tyres”, ISO, Geneva, Switzerland (2017).
7 ISO/TS 11819-3, “Acoustics — Measurement of the influence of road surfaces on traffic
noise Part 3: Reference tyres”, ISO, Geneva, Switzerland (2017).
8 W. Schwanen, H.M. van Leeuwen and A.A.A. Peeters, „Acoustic optimization tool RE3:
Measurement data Kloosterzande test track”, Report M+P.DWW.06.04.8, Vught, the
Netherlands (2007).
9 E. Bühlmann and T. Ziegler, “Temperature effects on tyre/road noise measurements”, Proc.
INTER-NOISE 11, (2011).
10 E. Bühlmann E., S. Egger, P. Mioduszevski and U. Sandberg, “An in-depth look at the tire
rubber hardness influence on tire/road noise measurements”, Proc INTER-NOISE 18,
Chicago (2018).
11 U. Sandberg and J.A. Ejsmont, “Tyre/Road Noise Reference Book”, Informex, Kisa, Sweden
(2001).
12 D. Lodico and P. Donavan, “Evaluation of temperature effects for on-board sound intensity
(OBSI) measurements,” TRB 2012 Annual Meeting (2012).
13 M. Bueno, J. Luong, U. Viñuela, F. Teran and S.E. Paje, „Pavement temperature influence
on close proximity tire/road noise,“ Applied Acoustics 73, 829-835 (2011).
14 F. Anfosso-Lédée and Y. Pichaud, “Temperature Effect on Tyre-Road Noise,“ Applied
Acoustics 68 (2007).
15 A. Kuijpers, “Temperature Effects on Tyre/Road Noise and on Tyre Stiffness,” M+P bv
memo to ISO Working Group 27, M+P.WG27/2002/2/ak (2002).
16 E. Bühlmann, U. Sandberg, P. Mioduszewski, Speed dependency of temperature effects on
road traffic noise, Proc INTER-NOISE 15, San Francisco (2015).
17 E. Bühlmann, S. Schulze and T. Ziegler, “Ageing of the new CPX reference tyres during a
measurement season”. Proc INTER-NOISE 13, Innsbruck, Austria (2013).
18 E. Bühlmann, P. Mioduszewski, U. Sandberg, „An in-depth look at the tire rubber hardness
influence on tire/road noise measurements”, Proc. INTER-NOISE 18, Chicago (2018).
... The newest standard corrects for slight deviations in driving speed, air temperature during the measurements, and rubber hardness of the test tire. Bühlmann showed in 2019 (Bühlmann, 2019) that after these corrections, repeated measures at the street granularity, taken at different days were well within one dB of each other and that the correlation between CPX and SPB is 0.87 for overall noise levels on roads with a range of 10 dBA in CPX levels. However in the spectral domain, frequency bands lower than or equal to 630 Hz show a correlation with SPB well below 0.8. ...
... Several things could have happened: the roads have degraded, the road has been resurfaced, precise driving track was slightly different on worn roads, or calibration of the CPX device was different in 2017. The expected uncertainty on repeated CPX measurements itself is expected to be within 1 dB for overall levels (Bühlmann, 2019). But at particular frequencies deviations e.g. between measurement devices may indeed be up to a few dB (Vieira and Sandberg, 2019). ...
Article
Currently, municipalities assess rolling noise on road surfaces using Close-Proximity measurements (CPX). To avoid these labor-intensive measurements, an opportunistic approach based on commodity sensors in a fleet of cars, is proposed. Blind sensor calibration eliminates the effect of measurement vehicle and varying observation conditions. Calibration relies on spatial coherence: modifiers and confounders do not interact strongly with location while the quantity of interest depends on location and less on measurement vehicle. Generalized additive speed models, car offset and de-noising autoencoders (DAE) were investigated. DAE achieves prominent results: (1) ratio of variability of measurements at a single location to the variability of measurements over all locations increases, (2) convergence of mean measurement at a location is faster, and (3) seasonal effects are eliminated. Finally, although the proposed method includes a diversity of tires, below 1600 Hz its results differ from CPX less than the difference between bi-annually repeated CPX measurements.
... Some researchers declare good correlation between CPX and other methods [28] with minor divergences. There are, however, researches questioning repeatability of the method due to many factors affecting the method. ...
Article
According to the valid standard, Close Proximity (CPX) method intends to evaluate the influence of road surfaces on traffic noise. Measurements may be carried out with the use of a self-powered vehicle or a special test trailer equipped with testing tyre towed by another vehicle. Two different testing devices took part in the research organized in Poland in order to determine the ranking of road surfaces in terms of acoustic parameters. Two tests (in the year 2018 and 2019) were carried out on 6 different road sections with different wearing courses. Road surfaces were ranked, which enabled comparison of the two measuring systems, based on the same standard. Obtained results revealed some differences within measured values ranging from about −1 dB to + 2 dB, which led to ranking changes concerning the best road surface in terms of limitation of traffic noise.
Article
Full-text available
Low-noise surfaces have become a common mitigation action in the last decade, so much so that different methods for feature extraction have been established to evaluate their efficacy. Among these, the Close Proximity Index (CPX) evaluates the noise emissions by means of multiple runs at different speeds performed with a vehicle equipped with a reference tire and with acoustic sensors close to the wheel. However, signals acquired with CPX make it source oriented, and the analysis does not consider the real traffic flow of the studied site for a receiver-oriented approach. These aspects are remedied by Statistical Pass-By (SPB), a method based on sensor feature extraction with live detection of events; noise and speed acquisitions are performed at the roadside in real case scenarios. Unfortunately, the specific SPB requirements for its measurement setup do not allow an evaluation in urban context unless a special setup is used, but this may alter the acoustical context in which the measurement was performed. The present paper illustrates the testing and validation of a method named Urban Pass-By (U-SPB), developed during the LIFE NEREiDE project. U-SPB originates from standard SPB, exploits unattended measurements and develops an in-lab feature detection and extraction procedure. The U-SPB extends the evaluation in terms of before/after data comparison of the efficiency of low-noise laying in an urban context while combining the estimation of long-term noise levels and traffic parameters for other environmental noise purposes, such as noise mapping and action planning.
Article
Full-text available
Shore hardness measurements subject rubber bodies to standardized indentations. While easily performed, the measurements are subject to uncertainties. Although typical variances for Shore hardness are found in standards, operator and instrument effects are not well described, requiring statistically designed experiments to estimate effects and variance components. This paper focuses on uncertainty in Shore A hardness measurements of tyre tread elements and quantifies operator and instrument effects. Evaluation of uncertainty of Shore A measurements were performed on tyres under controlled conditions using three instruments, two tyres and five operators. Results show that the operator variance component and instrument effects are larger than the reference variance contribution in ISO 11819-3:2017. The interaction between operator and instrument is estimated to be the largest source of variation, while operator and instrument main effects are of similar size as the error component. Recommendations to reduce uncertainties include ignoring instantaneous values and requiring an instrument stand.
Conference Paper
Full-text available
When assessing the acoustic quality of a road pavement with the close-proximity (CPX) or the on-board sound intensity (OBSI) method, the rubber hardness of the reference tire substantially affects the measurement. Practical experience shows that measurement tires can get significantly harder within a single measurement season. This is why one would like to normalize measurements to a reference rubber hardness. The recently published technical specification defining the reference tires for CPX measurements (ISO/TS 11819-3), therefore, includes a new correction for tire rubber hardness. Early experiences with this new correction procedure raised questions about its accuracy. This paper takes an in-depth look on the influence of tire rubber hardness on CPX measurements for both reference tires P1 and H1. It analyses existing and new data and summarizes the research from several scientific contributions on this topic. It provides evidence that the effect of rubber hardness is tire specific and that separate correction factors for the P1 and the H1 tires lead to accuracy gains and improved repeatability and reproducibility of the method. The study concludes by proposing a revised tire-specific approach for the tire rubber hardness correction of CPX measurement results.
Conference Paper
Full-text available
Recently, low-noise road surfaces have become one of the preferred measures to reduce excessive traffic noise on the communal, cantonal and national road networks in Switzerland. This trend has been accelerated by a policy and subsidy scheme change as well as by new research outcomes with promising results in testing of a new generation of semi-dense asphalts with increased technical durability and longer-lasting noise reduction. Two main approaches were used by road owners for selection and quality control of low-noise road surfaces: firstly, by using a national standard which offers different options regarding chipping size and void content and, secondly, by relying upon the innovative capacity of road construction companies in developing products that correspond to a set of acoustical and technical requirements. This paper provides an overview of the challenges; the consequent research and the developed solutions which have proven to be critical milestones in the process of establishing low-noise road surfaces as an effective and durable noise abatement measure on Swiss roads. The construction of more than 1000 high quality low-noise road surfaces has resulted in the protection of a large number of people from excessive noise. PACS no. 43.50.Lj, 43.50.Gf
Article
Full-text available
Speed reductions and low noise road surfaces often represent the only option to reduce excessive traffic noise at the source. Noise abatement projects in urban areas therefore increasingly focus on the introduction of limiting speed to 30 km/h. However, existing noise emission models are commonly not designed for such low speeds and the peculiarities of a 30 km/h speed limit. Hence, the basis for a reliable prediction of the noise reduction by traffic calming measures for a 30 km/h speed limit is currently missing. The research project VSS 2012/214 provides an up-to-date basis for more reliable predic-tions on the noise reduction potential at low speeds, specifically at 30 km/h speed limit. Noise emissions were systematically assessed for different driving behaviours (gear se-lection, discontinuous driving and driving style) for a representative and up-to-date vehi-cle fleet during a comprehensive measurement campaign. The acquired data were further transferred into two separate emission approaches for constant and accelerated driving behaviour. These emission approaches were combined with a statistical survey on the actual driving behaviour at representative 30 km/h speed limit situations and an adapted emission approach for heavy vehicles from the European noise emission model CNOS-SOS and transferred into a source approach. The source approach allows an evaluation of 30 km/h speed limit situations regarding their noise reduction with a good reliability. The noise modelling results from this source formulation show that substantial noise re-ductions can be achieved by introducing speed limits of 30 km/h. Noise levels (Leq) can be reduced between approx. 2 dB and 4.5 dB, depending on the effective driven speed, the proportion of heavy vehicles and the road surface. Using an up-to-date and repre-sentative vehicle fleet as well as considering driving behaviour is also of great im-portance. Additionally, the results show a crucial dependency of the total noise reduction on the situation type and the type of the installed traffic calming measure. The situations investigated show that it is possible to achieve substantial speed reductions and a con-siderable reduction of noise levels, even without substantial road redesign. The effective speed, the proportion of heavy vehicles and the acoustic state of the road surface have been identified as the main sources for the variability in the total noise effect observed in various case studies by modelling the noise effect for different 30 km/h speed limit situations. The effective speed reduction is defined by the difference between the real driven speed in the initial and the target situation. While investigating the noise effect of 30 km/h speed limit situations, the effective speed reduction constitutes a crucial parameter for the noise effect. If the implemented traffic calming measures do not lead towards more discontinuous driving behaviour or towards driving in smaller gears, speed reductions of 10 km/h have been shown to be sufficient to realise substantial noise reductions. It is, however, necessary to be aware of the fact that the acoustic effect of a 30 km/h speed limit decreases with an increasing proportion of heavy vehicles. When considering proportions of heavy vehicles of more than 15%, a 30 km/h speed limit will generally only provide a small acoustic improvement. The acoustic state of the road surface also has a crucial influence on the noise effect of a 30 km/h speed limit. Generally, the louder the road surface or rather the stronger it con-tributes to the rolling noise formation, the more distinctive is the potential noise reduction of a 30 km/h speed limit. Depending on their specific noise parameters, low noise road surfaces can cause an additional noise effect of up to -2 dB in speed limit 30 situations. Hence, a combination of a speed limit of 30 km/h with a low noise road surface can be advantageous in certain situations. Generally, the acoustic effectiveness of road surfaces is expected to be higher at low proportions of heavy vehicles. Recommendations for further research is to investigate the noise effect of 30 km/h speed limits at high proportions of trucks or buses, the noise effect of traffic calming measures, the effect of combining low noise pavements and 30 km/h speed limits and the noise effect of 30 km/h speed limits on road sections with slopes (more than 6%), in order to further increase the prediction reliability
Conference Paper
Full-text available
Tyre/road noise measurements are significantly influenced by temperature. Understanding this effect is crucial to ensure repeatable and reproducible tyre/road noise measurements through the application of temperature correction procedures. These procedures currently lack accuracy and need to be reconsidered. This paper analyses how temperature affects CPX (close proximity) measurements using the reference tyres SRTT and Avon AV4. A series of consecutive noise measurements were performed on clear days on various road surfaces (12 road surface types of different ages, with speeds of 80km/h and 50km/h, air temperature range 10–30°C). The " surface-least-affected " air temperature (height 150 cm) was found to be the most suitable for correction of temperature effects when using single correction factors for situations with varying solar radiation and road surface colour. Significant temperature effects were found for both tyres on all assessed road surfaces. The greatest effects were found for high frequencies (1600–5000 Hz) in agreement with previous studies. This study, however, also revealed considerable temperature effects (-0.09 dB(A)/°C on average for dense asphalts) for mid frequencies (800–1250 Hz), a range which accounts for more than half of the overall effect. The average overall noise-temperature slopes on dense asphalts were found to be-0.10 dB(A)/°C (SRTT) and-0.11 dB(A)/°C (Avon AV4). The temperature effects found on cement concrete and porous asphalts were somewhat lower, though still substantial. Furthermore, this study highlights the need for a semi-generic and spectral approach when correcting temperature effects.
Article
The aim of this work is to analyze the influence of the surface temperature on the acoustical behaviour of a semidense asphalt pavement located in an urban area. The sound levels emitted by the interaction between a reference tire and the asphalt pavement at different surface temperatures were measured with the trailer Tiresonic Mk4 LA2IC-UCLM rolling at a speed of 50km/h. The analysis of the results shows that increasing pavement temperature leads to a reduction in the close proximity sound levels assessed at a rate of 0.06dB(A)/°C. Moreover, spectral analysis confirms that both the mechanisms associated with vibration and impacts and those related to the friction and adhesion between tire and pavement in the contact patch could be affected by the variation of the surface temperature.
Article
Tyre–road noise emission decreases when the outdoor temperature increases, with a variation that can exceed −0.1dB(A)/°C. This effect depends on tyre–road combination, but semi-generic corrections can improve the accuracy of tyre–road noise measurements. In this paper, the variation of pass-by noise level of a passenger car at 90km/h with temperature is investigated, on seven types of road surfaces, under different temperature conditions. A good correlation between air, road surface and tyre temperature is outlined. A linear relationship between noise level and air temperature variations is observed for bituminous pavements, of about −0.1dB(A)/°C, but reduced to −0.06dB(A)/°C for pavements having porosity. No temperature effect is observed on cement concrete pavements. A spectral analysis shows that the temperature effect is highest in low and high frequency range, what can be explained by generating mechanisms rather than propagation.
Use of noise reducing pavements -European experience
  • H Bendtsen
  • J Kragh
  • E Nielsen
H. Bendtsen, J. Kragh and E. Nielsen, "Use of noise reducing pavements -European experience", Hedehusenen, Denmark (2008).
Acoustic optimization tool -RE3: Measurement data Kloosterzande test track
  • W Schwanen
  • H M Van Leeuwen
  • A A A Peeters
W. Schwanen, H.M. van Leeuwen and A.A.A. Peeters, "Acoustic optimization tool -RE3: Measurement data Kloosterzande test track", Report M+P.DWW.06.04.8, Vught, the Netherlands (2007).
An in-depth look at the tire rubber hardness influence on tire/road noise measurements
  • E E Bühlmann
  • S Egger
  • P Mioduszevski
  • U Sandberg
E. Bühlmann E., S. Egger, P. Mioduszevski and U. Sandberg, "An in-depth look at the tire rubber hardness influence on tire/road noise measurements", Proc INTER-NOISE 18, Chicago (2018).
Tyre/Road Noise Reference Book
  • U Sandberg
  • J A Ejsmont
U. Sandberg and J.A. Ejsmont, "Tyre/Road Noise Reference Book", Informex, Kisa, Sweden (2001).