In this paper, a numerical approach to simulate the flow-induced far-field noise from a high speed train is investigated, using a digital train design specially developed for this purpose. The focus is made on the first bogie and pantograph contributions for a train speed of 320 kph. The numerical method uses a Lattice Boltzmann Method (LBM) based CFD/CAA solver combined with a Ffowcs, Williams and Hawkings (FW-H) solver for noise propagation to the far-field. A grid resolution study is performed for both of these important noise source regions, and the use of symmetry plane to reduce simulation cost for the bogie case setup is shown to be acceptable. Both the porous and solid FW-H formulations are studied and the effect of the inputs surface location is investigated. For pantograph predictions, the porous and solid formulations favorably compare both to each other and to the direct noise prediction from the LBM flow simulation. For the bogie predictions, the porous formulation matches the direct noise prediction, while the surface formulation is less accurate due to the screening effect of the bogie skirts. The strong agreement between FW-H predictions and direct simulation results at mid field locations provide a compelling consistency check supporting the accuracy of the numerical approach. Analyses of the detailed transient flow information are used to propose some design modifications for noise reduction in each area. Overall noise contributions due to the pantograph and the first bogie are presented for the far field locations used in the technical specification for interoperability (TSI) certification process. The results presented here tend to indicate that the simulations can be used as a design tool both to assess the far-field noise levels, important to meeting regulatory requirements, and to explore noise reduction concepts for key noise sources. Comparisons to test data would improve further the confidence on the absolute predicted values, but obtaining such data in the form of precise and reliable measurements is a challenging task. The compute time for each simulation is about one day for pantograph noise and about 2.5 days for first bogie noise, assuming a 532 cores Linux machine (at today's typical processing speed). This amount of compute power allows the predictions to reach a frequency of 5000 Hz for the pantograph and 2000 - 3000 Hz for the first bogie. Further studies concerning design modifications for bogie and pantograph, as well as interior cabin and intercoach noise, will be presented in the future.


    Zugriff

    Zugriff über TIB

    Verfügbarkeit in meiner Bibliothek prüfen


    Exportieren, teilen und zitieren



    Titel :

    High speed train aeroacoustic solution for community noise using lattice Boltzmann Method


    Beteiligte:
    Meskine, M. (Autor:in) / Perot, F. (Autor:in) / Kim, M.S. (Autor:in) / Senthooran, S. (Autor:in) / Freed, D.M. (Autor:in) / Sugiyama, Z. (Autor:in) / Polidoro, F. (Autor:in) / Gautier, S. (Autor:in)


    Erscheinungsdatum :

    2012


    Format / Umfang :

    11 Seiten, 27 Bilder, 2 Tabellen, 23 Quellen


    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Datenträger


    Sprache :

    Englisch




    Accuracy of Lattice Boltzmann Method for Aeroacoustic Simulations

    Marié, Simon / Ricot, Denis / Sagaut, Pierre | AIAA | 2007


    Aeroacoustic Analysis of a High Lift Trapezoidal Wing using Lattice Boltzmann Method

    Satti, Rajani / Li, Yanbing / Shock, Richard et al. | AIAA | 2008


    Aeroacoustic investigation of automotive engine cooling modules using the Lattice-Boltzmann Method

    Tebib, Safouane / Ballapur Jayasimha, Athreya / Moreau, Stephane et al. | AIAA | 2022



    Wind Tunnel Tests on the Control of Aeroacoustic Noise from High Speed Train

    Yamazaki, N. / Takaishi, T. / Toyooka, M. et al. | Springer Verlag | 2008