Sound emissions of an automotive engine cooling system are studied using both single-microphone directivity measurements and a rotating beamforming technique. These measurements provide reference acoustic data on such a system and some new understanding of the effect that the radiator induces on the distribution of sound sources. Indeed, the beamforming results indicate that, above the frequency limit allowed by the Rayleigh criterion, it is possible to localize and quantify the noise sources even through the heat-exchanger core. Moreover, for the investigated operating points along the fan performance curve, the sources are always distributed at the tip of the blades and, in particular, at the leading edge. The present evidence, confirmed by the similar trends of the frequency spectra with and without the heat exchanger, leads to the conclusion that the dominant sound mechanism is the turbulence-interaction noise. Nevertheless, this turbulence is produced within the gap between the fan ring and its casing rather than generated by the radiator core. The latter appears to induce negligible acoustic transmission losses but, more significantly, is found to have a minimal influence on the aerodynamic modification of sound sources for all the analyzed operating conditions.

Aerodynamic noise emitted by low-speed axial fans has been receiving increasing attention in various sectors of high societal impact, such as automotive and HVAC systems. In this framework, turbulence interaction, flow non-uniformities, trailing-edge boundary layer fluctuations and blade-tip leakages are different mechanisms generating aeroacoustic sources on the rotating blades and contributing to the overall emitted sound. An accurate localization of the sound sources on the surface of the blade is instrumental in separating and isolating these contributions and, therefore, in designing novel sound mitigation concepts. The main objective of this paper is to present an inexpensive and efficient way to isolate and quantify the noise generating mechanisms on rotating blades by means of a irregularly shaped microphone array. The technique is based on ROtating Source Identifier (ROSI) and has been implemented and validated at the von Karman Institute for Fluid Dynamics (VKI). Simulated benchmark datasets that refer to rotating point sources emitting white noise have been considered for the validation of the method. The accuracy in the source localization and in the source strength reconstruction has been evaluated for a fixed and a variable angular rate. Moreover, the algorithm implementation has been parallelized with the purpose of reducing its computational time, which represents the main drawback of ROSI. Finally, the developed technique has been applied to measure the noise sources generated by a forward-skewed subsonic axial fan operated at maximum efficiency. In this case, it has been possible to successfully localize and characterize the major noise sources on the blades. Although further investigation will be necessary to gain better insight into the topic, the present work constitutes an important step for a better understanding of the physical phenomena occurring in the noise generation of an axial fan.