EN 13848-2-2020 pdf download.Railway applications – Track – Track geometry quality – Part 2: Measuring systems – Track recording vehicles.
Due to the risk of high dynamic track responses, specific track sections may be excluded from the validation as for example:
— class E quality sections for the same speed range;
— track sections that do not conform to plain line (e.g. sidings, switches and crossings…);
— track sections that include degraded components such as loose fasteners.
The total length of the test sections used for validation should not be less than 5 km (typical length:
10 km). For speeds lower than or equal to 40 km/h a shorter total length of the test sections e.g. 1 km may be used.
A reference track may be used to ensure a better comparability between systems and over time. Data defining the characteristics, such as curve radii, and geometric quality of the track used for the test runs shall be provided with the test reporl
6.3.3.4 Comparison of different runs
6.3.3.4.1 Overview
Two types of analysis can be carried out:
comparison of runs of the same recording system;
— comparison of runs of a different recording system.
The statistical analysis of parameter data shall be used to compare recording runs (see 6.3.3.4.3).
The frequency analysis should be used (see 6.3.3.4.4).
Additionally, the uncertainty analysis can be performed (see 6.3.3.4.5).
6.3.3.4.2 Synchronization
To get reliable results when comparing parameters, data synchronization is necessary. Because of the effects of drift on the measured distance, this drift should be computed and compensated on short sections of e.g. 200 m length by shifting the signal. All track geometry parameters shall be compensated by applying the same shift.
Other treatments such as resampling or interpolation may be performed.
6.3.3.4.3 Statistical analysis of parameter data
6.3.3.4.4 Frequency analysis of parameter data
Another way to compare two runs is to use frequency analysis such as transfer and coherence functions. A brief description of these functions is given in Annex A.
With this method it is possible to check the similarity in frequency domain of two measurements of the same track section. When applying this method a total length of the test sections of at least 5 km should be used.
The transfer function indicates the distortions existing in frequency domain, and the coherence function, the degree of reliability for the calculation made on transfer function.
The modulus of the transfer function should be as close as possible to unity and the phase as close as possible to zero for the wavelength ranges specified in EN 13848-1.
The fixed relationship between the sample size and the wavelength range (Dl, D2, D3) in which the frequency analysis is made should be considered to determine the minimum length of the test section (see Annex A).
6.3.3.4.5 Measurement uncertainty analysis
The estimation of measurement uncertainty can be used to evaluate the quality of the measurement
system and to determine lithe measurement process is lit for its intended use.
A suitable method and limit values for measurement uncertainty are proposed in Annex D.
6.3.3.5 Cross check
Cross checks can be performed in the frequency and in the space domain.
A method for the frequency domain is described in A.3.2 and the results shall comply with the limit values given in C.4.
A method for the space domain is described in Annex E. Limit values are still an open point due to the lack of experience.
Examples of linked parameters are longitudinal level of each rail versus cross level or alignment of each rail versus gauge. Correct filtering of all input parameters has to be taken into account.
6.3.4 Methodology and frequency of the validation tests
6.3.4.1 Overview
Depending on the circumstances different validation tests shall be carried out. They are described in the following sections.
6.3.4.2 Validation of a new system.EN 13848-2-2020 pdf download.