Studies dedicated to the determination of acoustic characteristics of an automotive cooling package are presented. A shrouded subsonic axial fan is mounted in a wall separating an anechoic- and a reverberation room. This enables a unique separation of the up- and downstream sound fields. Microphone measurements were acquired of the radiated sound as a function of rotational speed, fan type and components included in the cooling module. The aim of the present work is to investigate the effect of a closely mounted radiator upstream of the impeller on the SPL spectral distribution. Upon examination of the SPL spectral shape, features linked specifically to the source and system are revealed. The properties of a reverberant sound field combined with the method of spectral decomposition permit an estimation of the source spectral distribution and the acoustic transfer response, respectively. Additionally, purely intrinsic acoustic properties of the radiator are scrutinized by standardized ISO methods. A new methodology comprising a dipole sound source is adopted to circumvent limitation of transmission loss measurement in the low frequency range. The sound attenuation caused by the radiator alone was found to be negligible.

To obtain realistic noise characteristics from CAA studies of subsonic fans, it is important to prescribe properly constructed turbulent inflow statistics. This is frequently omitted; instead it is assumed that the stochastic characteristics of turbulence, absent at the initial stage, progressively develops as the rotor inflicts the flow field over time and hence that the sound generating mechanism governed by surface pressure fluctuations are asymptotically accounted for. That assumption violates the actual interplay taking place between an ingested flow field and the surface pressure fluctuations exerted by the blades producing noise. The aim of the present study is to examine the coupling effect between synthetically ingested turbulence to sound produced from a subsonic ducted fan. The steady state inflow parameters are mapped from a precursor RANS simulation onto the inflow boundaries of a reduced domain to limit the computational cost. The flow field is resolved utilizing IDDES for turbulence handling and the computational domains are configured for both a single blade and a circumferential complete five bladed fan. The results clearly reveal the limitations of restricting the computational analysis to a single blade. Additionally, a separate investigation of the upstream inlet section shows that the deterministic flow structures generated at the inlet plane are selfsustained within the inlet section. The outcome stresses the importance of incorporating correlated inflow statistics for turbomachinery noise studies. Moreover, the acoustic analogy formulated by Ffowcs -Williams and Hawkings is employed to study the low frequency spectral distribution. Previously conducted measurements are used for validation of both the flow field statistics and the far-field sound field.

A study has been conducted to numerically examine the noise generation mechanisms associated witha low-speed fan operating in a turbulent-rich inflow. Although, the turbulent content in the upstreamflow is an important aspect of the noise characteristics associated with subsonic fans, it is often omittedin numerical investigation. In this paper the turbulent structures are generated by the synthetic eddymethod (SEM) which ensures that the vortical structures imposed on the inflow boundary are temporallyand spatially correlated. The flow field is resolved using the Improved Delayed Detached Eddy Simulation(IDDES), which enables in addition to the deterministic tonal components associated with the blade passingfrequency (BPF) and its harmonics, the continuous envelope of broadband noise associated with turbulenceto be scrutinized. In the present work, the noise produced from the tip leakage flow is suppressed by aslightly less resolved tip region. Hence the effect on the noise produced associated with both the inflowturbulence and the vortices in the tip region are examined. The sound pressure level (SPL) in the far-fieldis obtained from the Ffowcs–Williams and Hawkings acoustic analogy. Validation with experimental resultsof a similar setup is provided. The results presented accentuates the need for properly constructed inletboundary conditions where turbulent structures, either part of the inflow or developed in the tip region,reinforce the noise produced.

The characteristics of the noise radiated from a reduced automotive cooling module is numerically studied with focus on the interaction effects linked to the sound generation mechanisms and the acoustic scattering caused by the confined installation. The flow field is simulated by adopting the formulation of IDDES, which is a numerical technique that enables large scale structures to be resolved and the wall-bounded flow to be treated either in DDES or WMLES depending on the turbulent content within the boundary layer. By comparing the simulated fan performance to two sets of experimental data of a similar setup, the aerodynamic results obtained from IDDES are validated and conformed to the volumetric flow rate delivered for the pressure drop measured. The acoustic part of the study comprises evaluation of the sound source associated with the momentum distribution imposed on the surroundings at an interface slightly upstream the fan. At the microphone positions upstream the installation, the SPL falls within the SPL range measured and the acoustic power delivered by the fan conforms to the SWL obtained from the comparison method in the reverberation room. The system response function, estimated by subtracting the SWL for the free-field simulation from the SWL associated with the reduced automotive cooling modulemarks spectral humps at fixed frequencies, irrespectively of sound source. As such, the engineering approach to spectral decomposition method earlier published which, enables isolating the acoustical properties of the installation from the source, is validated and found to hold.

An automotive radiator, situated closely upstream of an operating fan, modifies the flow field impinging on the fan blades. This in turn is reflected in the noise produced. In addition, the operating condition is changed attributed to the intrinsic pressure loss of the radiator. Further, additional structural elements scatter the sound originating from the impeller due to extraneous or intrinsic sources imposed by the radiator in the system. This paper numerically examines these effects by aerodynamically and acoustically modeling of the radiator as a porous medium. The flow field is resolved utilizing the methodology of improved delayed detached eddy simulation (IDDES) and the sound in the far-field is computed by a finite element sound propagation solver given the velocity components on an interface defined as the sound source plane. The sound pressure level (SPL) is presented for different rotational speeds and microphone positions and is compared to the measured SPL of a similar setup. The acoustic properties of the radiator alone is related to the sound transmission behavior simulated and measured. The results presented are in accordance with the behavior acquired experimentally.