The environmental impact of the railways operations – the dependence of speed on the level of vibration emission

Railway transport plays a significant role in shaping the quality of the environment. Vibration from railways is a big concern for engineers. Because of its negative impact on the surrounding area, it is also an environmental problem. The paper deals with the dependence of speed on the level of vibration emission. It based on the experimental study, which was done during the passage of the Pendolino train (EMU 250). Accredited Laboratory of Structural Mechanics at Cracow University of Technology has performed in situ measurements. Speed of the train was varying between 80 and 250 km/h. The evaluated construction was located at distance of approximately 50 m from the railroad. Indicators of the perceptibility of vibration through constructions (WODB) were used to access the impact of vibrations on the building. The innovation of this publication is the application of multi-criteria decision analysis methods for assessing the impact of vibration. The ELECTRE (Elimination and Choice Expressing Reality) methods to evaluate the increase in vibration emission depending on the speed for each frequency band are used. The limits specified by Polish standards in any of the passages have not been exceeded. As proven in the paper, a train speed increase does not uniformly make the vibration higher in each frequency band.


Introduction
The issue of vibration emission is governed by several standards and laws (inter alia [1][2][3]). With the attention paid to this subject in international literature [4][5][6][7], it is clear that the vibration induced by human activity is an important consideration throughout the modern world. It is well known that vibration induced by trains has a negative impact on the nearby area [8][9][10][11]. The increases in train speed and the creation of new high-speed lines increases the size of this problem.
The aim of this paper is to implement an application of Multi-Criteria Decision Analysis (MCDA) method ELECTRE I to assess the impact of train speed to perceptibility of vibration through constructions. ELECTRE, which is a family of methods, is widely known and used in scientific work [12][13]. Nevertheless, the value of this work is the combination of WODB (indicators of the perceptibility of vibration through constructions) with the ELECTRE I method (Elimination and Choice Expressing Reality).

Background and measuring polygon
Before the authorisation for the use of the Pendolino trains (Electric Multiple Unit -EMUs 250), homologation and velocity tests were conducted in Poland.The new train was tested whilst crossing the Psary -Góra Włodowska section (a length of approximately 36km) located within railway line No.  [14][15][16].
To perform measurement, a PCB Piezotronics piezoelectric accelerometer and a SCADAS Mobile LMS International analyser were used. Accelerometers were fixed on the building foundation. The investigated residential building was 50m from track A (where Pendolino were tested). The surveyed constructionwas a typical residential masonry building.
The paper presents selected results of measurements of building vibration induced by train at speed varying between 80 and 250 km/h.The station measured two mutually orthogonal horizontal components x and y of vibration accelerations labelled P-01x and P-02y, respectively. Fig. 1. shows the measuring polygon and the schematic arrangement of the sensors.  Figure 2 shows the maximum values obtained from the time histories of the acceleration of horizontal components (P-01x and P-02y) of vibrations caused by passages of the Pendolino trains with speed 80-250 km/h. The growth curves (exponential function) of maximum acceleration of vibration versus velocity of the train were fitted. It describes the dependence of the maximum acceleration of train speed. The coefficients of determination R 2 for both functions is above 0.9. Based on only these results it can be stated that the vibrations are increasing with the speed increase.
Nevertheless, scales of dynamic influences (SDI scales), Polish code [17] as mandatory recognizes not maximum values of vibration. Dimensions of the examined construction classify it as being within the scale SDI -II. By analyzing the results obtained according to the guidelinesit was found that these vibrations are not harmful to the building considered as a whole structure. All acceleration values in one-third octave bands are located in the first zone of the SDI scale.

Analysis of combination of WODB with ELECTRE I method
The indicator (WODB) is the maximum value of the ratio of the vibration acceleration parameters in one-third octave bands to the acceleration corresponding to the lower limit consideration of the dynamic influence on buildings due to the SDI scales in the same frequency band. WODB directly displays how many times the limit consideration of dynamic influence on construction has been exceeded.The harmful of vibrations is considered in the 1/3 octave bands up to 100Hz.Therefore, there are twenty -one frequency bands. Each event (a single passage of a train) generates 21 indicators. Therefore, comparing two events is no longer as easy as for maximum acceleration value (amax). It requires the use of more complex methods, which provides sensible and straightforward rankings. The purpose of applying ELECTRE is to evaluate whether higher velocity always produces higher vibrations. It is compared events at seven different speeds. The criterion is only WODB value for twenty-one bands. Each band is considered equivalent, therefore the weight is not introduced. In addition, in any considered case threshold of vibration harmfulness is not exceeded (first zone of the SDI) resulting there had not to be implemented activities to reduce the level of induced vibrations.In order to emphasize the various dependencies, different thresholds were used in the range of 0.5-0.6. The resulting matrix and graph ranking are shown in Figures 3-4. The relationship between the different thresholds and the two sensors varies. However, repetitive dependencies could be noticed. A train ride at a speed of 180km/h does not affect the environment much more than one at a speed of 160km/h. For the sensor P-01x and at threshold 0.6 and 0.7 it is clearly seen, that we have speed bands within which speed increases will not produce significantly greater vibrations (e.g. 80-120km/h; 160-  v 80 120 160 180 200 230 250 v 80 120 160 180 200 230 250  80 1 1 1 1 1 1 1 80 1 1 1 1 1 1 1  120 0 1 1 1 1 1 1 120 0 1 1 1 1 1 1  160 0 0 1 0 1 1 1 160 0 0 1 0 1 1 1  180 0 0 1 1 1 1 1 180 0 0 1 1 1 1 1  200 0 0 0 0 1 1 1 200 0 0

Conclusion
This study deals with the negative impact on surrounding structures caused by the passage of the Pendolino trains. Based on the analysis, we can draw the following conclusions: • The limits specified by Polish standard in any of the occurrences of train passage were not exceeded.The train has been put into service. • Maximum values obtained from the time history of the acceleration (amax) shows the upward / downward trend of vibration but it is not an indicator recognized by the standard.Becauseit is only a single peak, it may mislead or underestimate the total vibration level. • Propagation of vibrations induced by trains with their effect on the structural response of construction is a complex phenomenon. It is difficult to assess the influence of vibration on buildings without in situ measurements. A train moving on a track at variable speed does not uniformly make the vibration higher in each frequency band. • Standard indicators can successfully be treated as a multi-criterion problem. The use of analyzes e.g. from the family of ELECTRE could supplement the obtained information and highlight the omitted phenomena and dependencies.