Noise immissions are composed of the direct sound and reflections at buildings and structures. The reflections are difficult to calculate and so they are often disregarded. Acoustic measurements on scale models are usually too expensive.
For this reason the purpose of the project is to establish rules for calculating and judging the reflections for typical situations.
The research has focused on the following practical objectives:
- Show influence of reflections for different types and dimensions of structures such as underpasses, galleries, bridges, etc. on the intensity and characteristics of the decisive noise pollution
- Development of models to quantify the reflection shares
- Development of evaluation criteria for the arrangement and design of measures incl. Cost / benefit analysis (in particular for the lining with absorbent materials). It should be ensured the sustainability the noise.
The results of the tests are documented and the resulting findings should be simple and practically oriented applicable and form the basis for a standard.
To process the following sub-steps:
- Literature Study: Further evaluation of domestic and foreign studies as well as guidelines and standards on this subject
- Typing of buildings and definition of typical situations. The range of situations, which occurs in practice to be imaged, as well as possible with a limited number model types.
- Technical measurement of model situations by reference and comparative situations / evaluation of the frequency curves
- Mathematical simulation and model building with suitable programs (eg Cadna, Finite Differences-Time Domain Model of EMPA)
- Analysis and comparison of the metrological and theoretical studies and evaluation of the different influencing parameters (location immission, dimensions, materials, frequency content, etc.)
- Creation of measures with cost / benefit analysis, conclusions
- Development of evaluation criteria for the quantification of reflections and the assignment of model types and the application of measures
As a supplement is to be considered that future renovations, and possibly even new construction measurements before (without absorbing measures) and subsequently are to be taken (with absorbent linings). These can respectively by the cantons. Their agents are established. These measurements must be carried out (for reasons of time) outside the present research, however, could be included in a later revision of the results.
The study of literature in this field has shown that for complex propagation conditions and multiple reflections the model with regular reflections is in a lot of cases not sufficient. There have been developed specific models with different source and reflection models.
For proving the calculation methods it has been built a real model of an undergrade crossing in the scale 1:32. The immissions have been measured in the model and calculated with different models: A model with regular reflections (for simple and multiple reflections), a model with consideration of the Fresnel effect, and a model with area sources. The comparison shows:
- The best accord with the measurement results for the regular reflections you get, if you calculate only with the first reflection.
- The differences between measurement and calculation go up to 2.4 dB. These differences can be explicated with the exact geometrical position of the noise propagation rays, partly also with the lack of multiple reflections.
They have been made also measurements and calculations for a partly absorbent structure. The results are:
- Minimum the half part of the lateral wall should be absorbent
- The absorbent part should be placed in the higher part of the lateral wall.
- Absorbent areas are more efficient there where the ramp is deep.
The comparison with the alternative models "Fresnel" and "area source" has shown:
- The results of both models are very exact. They are closer to the measurement values then the calculation results with regular reflections.
- The calculation with "Fresnel" give good results in situation with small reflecting areas, the calculation with area sources is then favorably, when multiple reflections can have an important influence on the result.
An extension of these results on other situations is for practical reasons restricted on 2 dimensions. There has been proved 8 typical situations (lateral walls, galleries bridges, walls at both sides). The immissions have been calculated for 3 levels over ground (2 m, 5 m and 8 m) for a point in 37.5 m distance to the street.
The base for the comparison have been calculations with a finite element model. The comparison with the results of calculations with regular reflections (simple and multiple reflections) and the area source model shows:
- There are big differences up to 6 dB.
- Not any of t