appendix a - Velocidade

32
The effect of speed on noise, vibration and emissions
from vehicles
May 1998
With regard to the speed influence, the major models are only moderately divergent. An
interesting comparison of 7 of the major models appears in van Leeuwen et al, (1996). In the
range 50-100 km/h, the maximum deviation between the models can be illustrated as
follows. If all models were normalised to give the same value at 50 km/h (as an example),
they differ approximately 2 dB for light and 3 dB for heavy vehicles at any other speed, i.e.
this is the maximum deviation due to the speed effect. The Nordic model is a little more
speed sensitive than most of the other models. It is estimated here that in a hypothetical case
of calculating the noise reduction for a speed reduction of 20 km/h, the choice of model
could make a difference of around 1 dB in the results. Since ± 1 dB would be the best
possible precision of these models, such a potential error may be accepted.
The models do not make a distinction between rural and urban cases. The distinction is
implicit through the difference in traffic speeds specified for the urban and rural settings. It is
assumed that in an urban arterial the traffic is often congested so that it flows at lower speed.
In practice, however, none of the models would pick up the real difference that probably
exists between such cases, i.e. for a given speed the urban arterial would give slightly (but
only slightly) higher noise level because it is likely that the flow is not as free as in the rural
case and drivers may use a lower gear setting and accelerate and brake more. Since, for the
purposes of this study we cannot quantify this hypothetical effect, we shall assume the two
cases as similar.
As mentioned, the Nordic model ”overestimates” the speed effect on noise in relation to the
average of the other models. The reason may be that the Nordic model is based on very
recent data, and at least some of the other models may be based on rather old data, from a
time when power unit noise was dominating under more circumstances than now and the
influence of speed on noise was lower than currently.
It may be that some of the other models underestimate the speed effect. The authors
therefore believe that the Nordic model is the most relevant one for today and the future, but
this is of course arguable.
In the Nordic model, the noise-speed relation is as follows (LAE is ”the sound exposure
level” which is normalised to a distance of 10 m from road centre, and which is proportional
to the equivalent noise level by a factor depending on traffic flow):
LAE = 73.5 + 25. log (V/50) for V ≥ 40 km/h for light vehicles
LAE = 71.1 for 30 ≤ V < 40 km/h for light vehicles
LAE = 80.5 + 30. log (V/50) for 50 ≤ V ≤ 90 km/h for heavy vehicles
LAE = 80.5 for 30 ≤ V ≤ 50 km/h for heavy vehicles
These equations in the Nordic model are presented in Fig. 6.
This model, with the equations above as well as other parts of it, has been used in the
calculations presented in Appendix E. The software has been produced by Trivector (1997).