Vehicle Noise Emission Model (Coordination by RWTÜV)

The vehicle noise emission model consists of two parts, part 1, which calculates the noise emission for each second and each vehicle of the traffic simulation model in terms of rolling noise, propulsion noise and total noise as LA and 3rd octave band levels. The structure of part 1 is shown in the flowchart beside.

The first calculation step is the determination of normalised engine speed and load. This calculation may require several loops depending on the outcome of the plausibility check.

The following step is the calculation of LA and 3rd octave band levels for each dataset. The result is a copy of the input data file supplemented by vehicle acceleration, normalised engine speed and load and the noise levels. The noise levels represent the location 7.5 m distance from the drive path in the middle of the vehicle at a height of 1.2 m above the road.

The results table defines the detailed input data for part 2. It can also be exported for further analysis of single vehicle data for any user defined purpose.

The drivetrain model has to transform vehicle speed and acceleration pattern into normalised engine speed and engine load pattern. The calculation of normalised engine speed pattern requires gearshift algorithms, the calculation of normalised engine load requires normalised full load power curves and running resistance curves for different vehicle classes. The full load power curves are provided by the vehicle manufacturers amongst the project partners. For cars, light duty vehicles and motorcycles the running resistance curves are derived from running resistance coefficient tables used for exhaust emission measurements on roller benches.

In part 2 the information about location and time for each vehicle is used to aggregate and summarise all vehicles passing a road network section, that could be treated as single emitter, for each hour of the day. This part includes transformation algorithms from L_PP‘ to Leq at the reference distance of a propagation model (e.g. HARMONOISE). The flowchart of the calculation process is shown beside.

For part 2 an option is foreseen but not yet realised: An aggregation tool for the derivation of simplified aggregated input data for propagation models (vehicle categories instead of single vehicles, time intervals of 15 minutes or 1 hour instead of 1 second, characteristic 3rd octave band spectra for vehicle categories).

Noise emission measurements are carried out in order to calibrate the propulsion noise modelling as well as the rolling noise modelling. For the vehicles these measurements concentrate on the propulsion noise contribution. In total about 42 cars and light duty vehicles, 11 trucks, 1 bus and 3 motorcycles will be measured. All vehicles represent state of the art.

Rolling noise tests are performed for identifying and quantifying the external noise and determining the effect on objective results and the dependence upon the particular road/tyre/vehicle ensemble. The tests are performed on 3 different test surfaces, at different speeds, with 4 car tyre sizes, 2 types for light duty vehicles and 3 truck tyre sizes. One truck tyre size will be measured on 2-axle, 3-axle and 5-axle combinations separately.