Author Topic: What is the best way of segmenting roads?  (Read 1211 times)

roger

  • Administrator
  • Full Member
  • *****
  • Posts: 156
What is the best way of segmenting roads?
« on: May 23, 2018, 04:28:06 PM »
'Calculation of Road Traffic Noise', CRTN, states in para 11:  that roads should be divided into ‘a small number of segments’ such that ‘within any one segment the variation is less than 2 dB(A)’.

NoiseMap has an automated segmentation feature which generates segments along the road alignment to fit the horizontal and vertical curvature of the road to within a selected tolerance.  We recommend that this should be 1 m horizontally and 0.5 m vertically.

This is a geometric fit, rather than an acoustical fit.  However, it should be sufficient to deal with any variations in sound level due to distance and gradient.

Screening is more difficult because this is dependent on the position of the receiver relative to the position of screening objects. NoiseMap deals with this by sub-dividing the primary segments according to the position of the screening objects: indeed, the sub-divided primary segments are then further sub-divided by secondary screening objects. This sub-division is a complex process that takes up most of the calculation time, but ensures that multiple barriers are correctly assessed even if the primary segments are very long.

It should be noted that CRTN advises that roads are divided into a small number of segments.  This is an important proviso: in particular, if you follow the 'TRL' conversion procedure from L10 to Leq,t, this can be affected by the noise level emitted by each segment: very short segments will emit very little energy individually and this can distort the conversion formula.

Accordingly, we would not advise on pre-segmenting into say 10-metre segments as this could contravene the first part of the CRTN advice about ‘a small number of segments’ and then cause issues with converting the results to Leq,t.  This is particularly the case when using conversion methods 1 and 2.