• DocumentCode
    2934541
  • Title

    Large-scale Simulations of Turbulence: HPC and Numerical Experiments

  • Author

    Schlatter, Philipp ; Malm, Johan ; Brethouwer, Geert ; Johansson, Arne V. ; Henningson, Dan S.

  • Author_Institution
    Linne FLOW Centre & Swedish e-Sci. Res. Centre (SeRC), KTH Mech., Stockholm, Sweden
  • fYear
    2011
  • fDate
    5-8 Dec. 2011
  • Firstpage
    319
  • Lastpage
    324
  • Abstract
    In this paper, we describe some of the methods used at KTH Mechanics in the field of turbulence simulations. High performance computing (HPC) resources, including the dedicated system "Ekman" with 10000 cores at KTH, are employed to perform some of the largest turbulence simulations with up to 10 billion grid points. The results are used to assess the fidelity of such types of simulations by comparing in detail to wind tunnel experiments, excellent agreement is obtained in general. Simulations can thus be considered numerical experiments, and are subject to the same scrutiny as "real" experimental data. Validated simulation data is extremely valuable as it may provide unprecedented insight into the turbulence dynamics. However, given the large size of the computations, the massively parallel simulation codes, post processing tools and storage solutions have to be specifically adapted, making turbulence simulations an interdisplinary area of e-Science.
  • Keywords
    computational fluid dynamics; parallel processing; turbulence; high performance computing; large-scale simulation; massively parallel simulation codes; post processing tools; storage solutions; turbulence dynamics; turbulence simulation; Computational modeling; Computers; Geometry; Laboratories; Mathematical model; Numerical models; Program processors; boundary layers; direct numerical simulation; turbulence; turbulent separation;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    E-Science (e-Science), 2011 IEEE 7th International Conference on
  • Conference_Location
    Stockholm
  • Print_ISBN
    978-1-4577-2163-2
  • Type

    conf

  • DOI
    10.1109/eScience.2011.51
  • Filename
    6123294