Free Field Measurements of Large Objects During Operation

Measurements of a Wind Turbine

Measuring the noise emissions of wind turbines during operation is no easy task. The turbines are 50 metres high and surrounded by a loud background. The Acoustic Camera StarAC Pro is used here to analyse wind turbine noise during operation under free-field conditions. For this purpose, a wind turbine was analysed at different wind speeds between 7 and 10 m/s. Noise emissions occurring at the rear of the wind turbine were identified and localised using the NoiseImage software.

System Characteristics

Microphone Array Star48 AC Pro

This array is a 48 channel measurement system designed for outdoor applications. The lightweight aluminium array body is designed as slim as possible to ensure easy handling and accurate microphone positioning. The included high end tripod allows set-up in almost any measurement environment imaginable.

Data Recorder mcdRec

This data recording device has been developed especially for use with the Acoustic Camera. It facilitates high scanning frequencies and high numbers of channels but it is lightweight and compact at the same time.

Software

NoiseImage 3

Power Supply

Mobile power supply SilmMPS for the Independent usage of the Acoustic Camera for up to 4 hours.

2D Outdoor Measurement

Application area

Acoustic analysis of wind turbines during operation. 

Measurement task

Outdoor mapping of a wind turbine, localization of sound in different frequency bands at different wind speeds. For this purpose, a wind turbine has been measured at different wind speeds between 7 and 10 m/s. The noise emissions created by the turbine have been identified and localized with the software NoiseImage.

Measurement object

Wind power turbine (3 rotor blades, performance under 1 MW, height of mast 50 m, blade length 20 m).

Measurement set up

The measurements have been carried out with the microphone array Star 48 and a 48 Channels Data recorder connected to a mobile power supply with up to 4 hours of autonomy.
During the measurement, a relatively loud background noise coming from the wind and other environmental sources were recorded. It was minimized with the help of a filter.

Results

The main goal of the analysis was to form a conclusion about the focus of the noise emissions on a wind turbine. First, results of the A-weighted data show the flow noise at the blade tips, which depends on wind direction and wind turbine location in association with the position of the Acoustic Camera. For the measurement, the Acoustic Camera has been positioned slightly on the left behind the wind turbine. During the analysis of the blade noise, the acoustic focus was located to the left of the gear box which is the side with the shortest distances to the microphones.

For a comprehensive analysis, a whole rotation of one blade has been taken as calculation base. For the analyses, the software module Spectral Photo 2D has been employed which offers a pre-calculated location of the emissions for any frequency. The analysis led to the insight that the point of emission strongly depends on the observed frequencies. 

For the third octave band of 400 Hz the emission can be located clearly at the root part of the blade while this blade is turning downwards located at 140 degrees.

Looking at the frequency band from 706 to 891 Hz, it becomes obvious that the blade tips and also the gear box are emitting this flow noise.

It became clear that the higher frequencies are mainly emitted at the gear box while at the same time the hissing noises become visible as more quiet sources on the blade tips.

For further analysis a narrow band analysis was conducted – also with the help of the function Spectral Photo 2D. For this reason, the noise origin of the single peaks in the spectrum was analyzed. It was evident that the main source for this sine wave noise is the gear box.

Due to these results, a focus has been put upon the gear box itself. To focus onto a specific part of the measurement object it is not necessary to conduct a new measurement. The part of interest can easily be chosen during post-processing. By this proceeding it becomes clear that even at great distances it is possible to trace back certain frequencies to different emission points. In order to analyze the parts of the gear box themselves it is advisable to conduct a more detailed analysis from a shorter distance, possibly on a test bench.