The impacts of five different turbulence models on the accuracy of computational aeroacoustic results for an airfoil and acoustic localization analysis for well suited model

Journal article (2014)

Authors

Daniel Saatchi K.N. Toosi University of Technology
Mani Fathali K.N. Toosi University of Technology
AR Khojasteh K.N. Toosi University of Technology
Affiliation
External organisation
More Info
expand_more

Published Date

2014

Language

English

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Affiliation

External organisation

Abstract

The research consists of two main sections; it initially focused on the accuracy of turbulence models to simulate the acoustic far-field domain for a NACA 0012 airfoil at Mach number of 0.209. Five different turbulence models were used such as realizable k-ε, SST k-ω, Transition-SST, Reynolds-Stress, and Spalart-Allmaras in unsteady flow at velocity of 71.3 m/s. It used the Ffowcs-William and Hawkings (FW-H) acoustic method to compare the one-third-octaveband of semi-logarithmic sound pressure level (SPL) vs. frequency within the range of 200 Hz to 20000 Hz with experimental data. Reasonably good agreement is observed that both realizable k- ε and Transition-SST are more accurate than others in prediction of peak of sound pressure level and the frequency of peak. In second section, in retrospect to considering accuracy of realizable k- ε model and its shorter runtime, the paper considered some analysis on the acoustic far-field domain to find the critical points of fluctuated pressure due to the retard time. For this reason, a need to find a way to input the location of receiver to compare its data with other located receivers was felt. Indeed, it defined two different major distributions, semi-circle and far-field linear distribution, that each one has nineteen receivers. These distributions have mathematical correlations derived in their own subsection. It also introduces a parameter, Average-Dependable SPL, in order to analyze the effect of location and retard time on the fluctuations. After specification of six critical points in schematic contours, it is developed to a third distribution that corresponds to the proportion of chord. Finally, it can be observed that majority of peaks occur close to the turbulent boundary layer of trailing edge, but the quantity of peaks near the leading edge is more substantial than trailing edge peaks.