COMPUTATIONAL AERO-ACOUSTICS FOR VEHICLE APPLICATIONS
|PhD student (2000-02):||
|PhD student (2002-2008):||
|Cooperation:||Volvo Cars Corp., ANSYS Fluent Sweden*|
|Sponsors:||Volvo Cars Corp., Vinnova, ANSYS Fluent|
|Start of project:||October 2000|
Aerodynamic noise is the dominating noise source at car speeds above 100 km/h. The major noise sources are noise generated by the flow around A-pillar, mirrors, wheel housings, ventilation channels and fans.
Aeroacoustics (AA) is an attribute that has a large impact on product image and is directly experienced by the customer. Aeroacoustic noise problems are traditionally solved by addition of damping material and increased weight. This is in direct conflict with the goal to reduce the fuel consumption. The present project aims at reducing aeroacoustic noise at the source and thus avoid increased weight.
The work in [1,2] focus mainly on the scalar wave equations pioneered by Lighthill. A thorough review, including the inherent limitations in each method, is conducted. The equation by Curle is identified as the one that meets the objectives of this study the best, and it is implemented and solved for an open cavity. Since no experimental data is available for this case, a Direct Simulation (two-dimensional), resolving both the flow and the acoustics, is performed.
The sound field computed by Curle's equation agrees well with the Direct Simulation, but the computational cost involved is considerably smaller. This makes Curle's equation, together with the simplifications introduced and validated in this thesis, suitable for engineering use. The acoustic noise generation is studied in detail. The wall pressure fluctuations are found to account for roughly 90% of the radiated sound intensity, and the viscous and entropy fluctuations are found to be negligible. The downstream region of the cavity is found to generate the most sound, a fact which can be used to explain the directivity of the radiated sound.
Currently the unsteady fluid flow in the 2D cavity is computed using an incompressible flow. The acoustic noise generation at walls is compared with the 2D Direct Simulation data. The agreement is found to be good. For low Mach number flows where walls are present the dominating source terms in Lighthill-Curle's equation are terms involving the wall pressure fluctuations. Since these terms are mainly affected by hydrodynamic phenomena one can assume that the sound sources can be captured in an incompressible flow field. A modified version of the Lighthill-Curle's analogy is applied to study the near field acoustics of a laminar flow past an open cavity at a Mach number of 0.15 with the length-to-depth ratio of L/D=4. The aim of the work is to study the differences in compressible and incompressible sources to Lighthill-Curle's equation and their influence on the radiated sound.
We are currently working on the flow around a generic mirror. The flow field is computed with DES and we're using STAR-CD. The turbulent instantaneous fluctuations obtained from these simulations are then imported to our acoustic code based on the Lighthill-Curle's analogy. With this code the radiated sound is computed.
This page, Computational Aero-Acoustics for Vehicle applications, should be part of a frames system at www.tfd.chalmers.se/~lada/projects/proind.html