ISO/TS 13473-4:2008 表面プロファイルを使用した舗装テクスチャの特性評価—パート4：表面プロファイルのスペクトル分析 | ページ 6

3.1

(texture) wavelength

λ

quantity describing the horizontal dimension of the amplitude variations of a surface profile

Note 1 to entry: (Texture) wavelength is normally expressed in metres (m) or millimetres (mm).

Note 2 to entry: Wavelength is a quantity commonly used and accepted in electrotechnical and signal processing vocabularies. Since many users of this Technical Specification may not be accustomed to using the term wavelength in pavement applications, and because electrical signals are often used in the analyses of road surface profiles, there is a possibility of confusion. Hence, the expression “texture wavelength” is preferred here to make a clear distinction in relation to other applications

Note 3 to entry: The profile may be considered as a stationary, random function of the distance along the surface. By means of a Fourier analysis, such a function may be mathematically represented as an infinite series of sinusoidal components of various frequencies (and wavelengths), each having a given amplitude and initial phase. For typical and continuous surface profiles, a profile analysed by its Fourier components contains a continuous distribution of wavelengths. The texture wavelength in ISO 13473 is the reciprocal of the spatial frequency, the unit of which is reciprocal metre (equivalent to cycles per metre). See also 3.14.

Note 4 to entry: The wavelengths may be represented physically as the various lengths of periodically repeated parts of the profile.

3.2

profile sampling

selection of representative parts of a road surface of which the profile will be measured

3.3

profilometer

device used for measuring the profile of a pavement surface

Note 1 to entry: Current designs of profilometers used in pavement engineering include, but are not limited to, sensors based on laser, light sectioning, needle tracer and ultrasonic technologies.

3.4

measurement speed

v

speed at which the profilometer sensor traverses the surface to be measured

Note 1 to entry: Measurement speed is normally expressed in kilometres per hour (km/h) or metres per second (m/s).

3.5

digital signal sampling

determination of discrete measurement values of a signal at regularly spaced data points (and the subsequent conversion of these values into digital code)

Note 1 to entry: In this generic definition of digital signal sampling, the regular spacing of the data points may be applied either in the time or in the spatial domain, depending on the domain (time or space) in which the signal is captured.

3.6

sampling interval

distance between two adjacent data points on the surface, which is equal to the measurement speed divided by the sampling frequency of the sensor

Note 1 to entry: Sampling interval is normally expressed in millimetres (mm).

3.7

profile measurement length

l_p

length of an uninterrupted profile measurement

Note 1 to entry: Profile measurement length is normally expressed in metres (m) or millimetres (mm).

3.8

repetition interval

r

distance between the beginning of two consecutive profile measurement lengths, the latter as defined in 3.7

Note 1 to entry: Repetition interval is normally expressed in metres (m).

3.9

evaluation length

l

length of a sample from a profile which has been or is to be analysed

Note 1 to entry: The evaluation length may or may not be equal to the profile measurement length (but never greater).

Note 2 to entry: Evaluation length is normally expressed in metres (m) or millimetres (mm).

3.10

drop-out

measured point (sample) on the profile which is recognized as invalid, and which is usually discarded in the subsequent data processing

3.11

drop-out rate

percentage (%) of measured points within the evaluation length which are recognized as being invalid

3.12

zero-mean, slope-suppressed profile curve

Z ( x )

profile curve, Z ( x ), for which the mean level of the profile over the evaluation length has been brought to zero and for which long-wavelength trends have been removed

Note 1 to entry: To obtain a profile curve useful for mathematical calculations, it is necessary to remove any slope or long-wavelength component (slope suppression), as well as to bring the mean level of the profile over the evaluation length to zero (offset suppression). This can be accomplished by subtracting a least-squares fit from the profile, see 9.2. The resulting mean line of the profile is then at zero level. See illustration in Figure 1.

3.13

surface profile spectrum

texture spectrum

unevenness spectrum

spectrum obtained when a profile curve has been analysed by either digital or analogue filtering techniques in order to determine the magnitude of its spectral components at different wavelengths (3.1) or spatial frequencies (3.14)

Note 1 to entry: A texture spectrum presents the magnitude of each spectral component as a function of either texture wavelength or spatial frequency.

3.14

spatial frequency

inverse of (texture) wavelength

Note 1 to entry: Spatial frequency is normally expressed in reciprocal metres (m⁻¹); see also 3.1, Note 3.

Note 2 to entry: The term “frequency” used in the time domain, more precisely “temporal frequency”, corresponds to “spatial frequency” in the space domain.

3.15

surface (texture) profile level

L_tx,λ

logarithmic transformation of an amplitude representation of a surface profile curve Z(x), the latter expressed as a root mean square value

Note 2 to entry: Octave-band and one-third-octave-band filters are specified in 4.4 of ISO 13473-2:2002.

Note 3 to entry: Texture amplitudes expressed as root-mean-square values, whether filtered or not, may have a range of several magnitudes, typically 10^-5 m to 10^-2 m. Spectral characterization of signals is used frequently in studies of acoustics, vibrations and electrotechnical engineering. In all those fields, it is most common to use logarithmic amplitude scales. The same approach is preferred in this part of ISO 13473.

Note 4 to entry: Texture profile levels in practical pavement engineering typically range from 20 dB to 80 dB with these definitions.

3.16

power spectral density

PSD

quantity expressing the power contained in a signal per unit frequency or per unit wavelength as a function of frequency or wavelength

Note 2 to entry: In the case of a Discrete Fourier Transform of a sampled signal, the PSD may be defined as the squared magnitude of the components of the Fourier series divided by the effective bandwidth of the (narrow) bands of the Fourier spectrum (see 9.4)

Note 3 to entry: In the case of spectral analysis of a pavement surface profile, the signal is not a function of time but of evaluation length l . The Power Spectral Density may then be given as a function of the spatial frequency or the (texture) wavelength and will be expressed in the unit m²/m⁻¹ = m³ or in the unit m²/mm, respectively.

Note 4 to entry: The word “Power” in this designation originates from electric and acoustic signal terminology where signals incorporate actual power and where the squared amplitude is a measure of this power.