Welcome to the new IOA website! Please reset your password to access your account.

Volume : 44

Part : 2

Proceedings of the Institute of Acoustics

 

 

Impact of railroad switches on rail noise exposure near stations

 

Anders Genell1, Swedish National Road and Transport Research Institute (VTI), Linköping, Sweden
Erik Nyberg2, Swedish National Road and Transport Research Institute (VTI), Linköping, Sweden

 

ABSTRACT

 

According to Common NOise aSSessment methOdS in EUrope (CNOSSOS-EU) in Annex II of Directive 2002/49/EC, noise from road, rail and airplane traffic, as well as noise from industries, shall be assessed using this common method. For railway noise in Sweden, noise assessment has previously been done using the Nordic Assessment Method for Train Noise, revised 1996 (NMT96). NMT96 includes a simple correction for rail joints of +3dB and for rail switches of +6dB. CNOSSOS-EU instead introduces a speed dependent correction based on a third octave band wavelength spectrum adding up to 20dB rolling noise energy in lower frequencies and down to -40 dB in higher frequencies. Measurements recently performed for two different rail switch types along the Swedish rail system indicate that the frequency distribution corresponds well to the CNOSSOS-EU correction for one type of rail switch but not for the other, and for the overall level difference the opposite is true. In order to investigate to what extent this deviation is affecting noise exposure an inventory of more than 12000 rail switches along the Swedish railroad network has been performed to identify what types are situated in densely populated areas such as railway stations.

 

1. INTRODUCTION

 

According to Common NOise aSSessment methOdS in EUrope (CNOSSOS-EU) in Annex II of Directive 2002/49/EC [1], noise from road, rail and airplane traffic, as well as noise from industries, shall be assessed using this common method. The railway noise source model in heavily influenced by the Harmonoise/IMAGINE noise prediction models [2][3] as well as by the de facto branch standard development software TWINS [4].

 

 

Figure 1: A flow-chart representation of the TWINS model. Adopted from [5].

 

Figure 1 depicts a flow chart representation of the TWINS model, and as input data at the top is wheel irregularities and rail irregularities. In CNOSSOS-EU [1] a similar approach is used for rolling noise on the basis of total roughness level LR,TOT, as described in Equation (1) where LR,TR is the roughness level of the track and LR,VEH is the roughness level of the wheel.

 

 

Rail switch impact noise is also represented by a total effective roughness level for a rail section covering 50m before and 50m after the rail position along the track, as represented in Equation (2), where a term LR,IMPACT is added to the wheel and track roughness.

 

 

Table G-4 of CNOSSOS-EU contains impact roughness level coefficients to be used if no other data is available. In order to investigate the validity of these coefficients for impact noise from Swedish rail switches, measurements for two different types of switches were performed for a number of train types. Figure 2 shows a graph of the rail roughness levels representing rail switch impact noise addition in CNOSSOS-EU.


 

Figure 2: Rail switch impact roughness level as given i table G-4 of CNOSSOS-EU.

 

2. RAIL SWITCHES

 

Figure 3 shows a principal drawing of a single point rail switch. The course of events when a train is passing a rail switch is well described in [6]: “When a wheel passes over the crossing in the facing move (from the switch panel towards the crossing panel) it will first encounter the wing rail. Due to the outwards deviation of the wing rail, the wheel‒rail contact point will move towards the outside of the wheel profile. For a typical conical wheel profile, the rolling radius will decrease and the wheel will move downwards unless the wing rail is elevated. The reduced rolling radius on the crossing side will induce a yawing motion of the wheelset towards the crossing. Due to the check rail, the lateral motion of the wheelset is restrained and wheel flange interference contact with the crossing nose is prevented. When the wheel reaches and makes contact with the crossing nose, the contact load is quickly transferred from the wing rail to the crossing nose.”

 

 

Figure 3: Principal drawing of a single point railway switch. The red parts are guide rails that through interaction with the wheel flange ensures that the train follows the correct lateral position through the switch.

 

In Sweden, work has been carried out since 2014 to introduce a newer type of rail switch, denoted 60E, to over time replace the older types SJ50, BV50 and UIC60. The 60E switch has a transition zone with such a slightly raised wing rail in order to reduce the impact forces when transitioning to the crossing nose [7]. This could theoretically also reduce the impact noise from the train passing through the switch.

 

Table 1: Types of switches and their prevalence in the Swedish railway system, and amount of nearby population. “EV”, “DKV” and “3V” denotes single, double or tree-way switch respectively.

 

 

A number of different types of switches are employed in Sweden. Table 1 shows the 10 most common switch types as well as the amount of people living within 300m from a switch and thus could be exposed to the rail switch pass impact noise. The rightmost two columns show total population within 300m of a switch of each type and the maximum population within 300m of a single switch of each type. As can be seen, SJ50 and UIC60 are the most common types, with the highest number of people withing 300m, whereas the newer 60E only represents 1/10 to 1/5 of the amount of people. Figure 4 shows an example of how the population around a single switch has been calculated, with population data available in 100m squares.


 

Figure 4: Example of population squares within a radius of a rail switch of 300m

 

4. MEASURED AND CALCULATED NOISE

 

Noise from two types of switches was measured in the vicinity of Gothenburg, Sweden. Two UIC60 switches and two 60E switches. Figure 5 shows the time history, using 1/10th second time weighting, for the a-weighted noise level from a train passing through an UIC60 switch. As can be seen, there are clear peaks for each passing axis in the switch. The peaks are much less prominent for the same train recorded ca 100m before reaching the switch. The peaks are of similar height however, indicating that rail is much smoother around the switch that on the track further away, but also that the impact does not add significantly to the overall a-weighted noise level.

 

 

Figure 5: Noise level time history of a train passing through a UIC60 switch

 

When comparing to the time history of a train passing through a rail switch of the 60E type, it can be seen that the latter shows an increase in a-weighted noise level of about 10 dB for the rail switch impact noise, indicating that the rail roughness likely is more similar along both the reference position, in this case situated ca 100m after the switch, and close to the switch (Figure 6).

 

 

Figure 6: Noise level time history of a train passing through a 60E switch.

 

There are other differences than the type of switch for the two cases, such as train type and speed, but the difference between the time histories for the two switches are relatively constant for all different types and speeds recorded at the different sites. In order to compare to the impact noise addition included in CNOSSOS-EU, the impact addition was calculated for five different train types and for a range of speeds. Figure 7 shows the calculated impact noise additions from CNOSSOS-EU together with the difference in measured a-weighted Sound Exposure Level (SEL) between rail switch passage noise and reference track passage noise. A few things can be noted in this figure. One thing is that for electrically driven freight trains (GTE) the impact noise from the 60E is basically similar to the reference passage level, indicating that the overall sound energy of an entire freight train passage is not determined by the switch impact noise. For the two passenger trains X31 and X60 passing through the UIC60 switch, there is a large spread in SEL difference, where for some passes the reference passage is louder than the switch impact, again indicating that the rail roughness close to the switch is lower that for the rail further away, but since some passes also show increased noise level from switch impact of up to about 7 dB, there could also be a large influence from wheel roughness from individual trains. For the two passenger train types passing the 60E switch there is less spread in the SEL differences, and the impact seems to add 5-10dB in a weighted SEL for the entire passage. None of the measured SEL differences reach as high as the added impact noise in CNOSSOS-EU however, except for the X2 passenger train at the top speed of 200 km/h. This suggests that CNOSSOS-EU overestimates impact noise for the Swedish railway system, although one should be careful not to draw to definite conclusions from the relative few measurements performed.


 

Figure 7: Differences between switches and reference track in comparison to the added impact noise in CNOSSOS-EU as a function of speed.

 

5. CONCLUSIONS

 

There seems to be a large difference between different types of switches when it comes to impact noise. Considering that the measurements presented here indicate that the older type does not add very much in the overall noise level, exchanging them to newer types could potentially increase noise exposure. Close to 1.5 million people live within 300m from a switch of either of the most common types of switches and if the impact noise from these would increase by 5-10 dB the negative health effects and annoyance is at risk to be severe. However, the is a chance that difference might be less severe between the types of switches, considering that the rail maintenance might differ between the tracks where the switches are situated. When comparing to the CNOSSOS-EU rail switch noise addition one can see that CNOSSOS-EU seems to overestimate overall noise level from rail switches except for the X2 train type passing through the 60E type rail switch. The impact noise in CNOSSOS-EU is modelled like an increased rail roughness for a distance of 100m, thus behaving like a line source compared to the measured impact noise point source, which might contribute to the overestimation of rail switch impact noise in CNOSSOS-EU. There is also a question of how much railway traffic is flowing through the different types of switches. Future work should include investigating rail roughness around different types of switches as well as what types of switches experience the most railway traffic in densely populated areas so that a weighted impact noise “hazard indicator” could be calculated as a guide for the continuing work with exchanging rail switches in the Swedish railway system, and so that a more suitable rail impact noise correction can be included in future noise exposure calculations.

 

6. ACKNOWLEDGEMENTS

 

The majority of the work involved in this paper was commissioned and financed by the Swedish Transport Administration.

 

7. REFERENCES

 

  1. Commission Directive (EU) 2015/996 of 19 May 2015 establishing common noise assessment methods according to Directive 2002/49/EC of the European Parliament and of the Council.
  2. Salomons, E. & Heimann, D., Description of the reference model, Deliverable 16 of the Harmonoise project, HAR29TR-041118-TNO10, 22 December 2004.
  3. Dittrich, M. G., IMAGINE railway noise source model, default source data and measurement protocol, Deliverable from WP6 of the IMAGINE project, IMA6TR-050912-TNO01, 12 September 2005.
  4. Thompson, D. J., Hemsworth, B. & Vincent, N., Experimental validation of the TWINS prediction program for rolling noise, part 1: description of the model and method. Journal of sound and vibration 193.1 (1996): 123-135.
  5. Ögen, M., Noise emissions from railway traffic, Technical Report, VTI rapport 559A, Linköping, Sweden, 2006. ISSN 0347-6030.
  6. Pålsson, B. A., Optimisation of Railway Switches and Crossings, Doctoral Thesis, Department of Applied Mechanics, Chalmers University of Technology, Gothenburg, Sweden 2014.
  7. Ebelin, J. & Elmström, M., Ny spårväxelstandard 60E - En jämförande studie av de äldre växelsortimenten SJ50, BV50, UIC60 och det nyinförda växelsortimentet 60E. Master’s Thesis LTH School of Engineering, Lund University, Sweden, 2014 (in Swedish).

 

1 anders.genell@vti.se

2 erik.nyberg@vti.se