Browsing by Author "Fu, Lin"

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    A short note on a 3D spectral analysis for turbulent flows on unstructured meshes
    (Elsevier, 2022-11-25) Tsoutsanis, Panagiotis; Nogueira, Xesús; Fu, Lin
    We propose two techniques for computing the energy spectra for 3D unstructured meshes that are consistent across different element types. These techniques can be particularly useful when assessing the dissipation characteristics and the suitability of several popular non-linear high-order methods for implicit large-eddy simulations (iLES). Numerical experiments demonstrate the performance of several element types for iLES of the Taylor-Green vortex, where a significantly different dissipation and dispersion mechanism for each element type is revealed. The energy spectra results are dependent on the technique selected for obtaining them, therefore an additional established technique from the literature is also included for comparison to further analyse their similarities and their differences. These techniques can be an integral tool for the tuning and calibration of non-linear high-order methods that can benefit both explicit and implicit large-eddy simulations (LES).
  • Thumbnail Image
    ItemOpen Access
    UCNS3D: An open-source high-order finite-volume unstructured CFD solver
    (Elsevier, 2022-06-21) Antoniadis, Antonis F.; Drikakis, Dimitris; Farmakis, Pericles S.; Fu, Lin; Kokkinakis, Ioannis; Nogueira, Xesús; Silva, Paulo A. S. F.; Skote, Martin; Titarev, Vladimir; Tsoutsanis, Panagiotis
    UCNS3D is an open-source computational solver for compressible flows on unstructured meshes. State-of-the-art high-order methods and their associated benefits can now be implemented for industrial-scale CFD problems due to the flexibility and highly-automated generation offered by unstructured meshes. We present the governing equations of the physical models employed in UCNS3D, and the numerical framework developed for their solution. The code has been designed so that extended to other systems of equations and numerical models is straightforward. The employed methods are validated towards a series of stringent well-established test problems against experimental or analytical solutions, where the full capabilities of UCNS3D in terms of applications spectrum, robustness, efficiency, and accuracy are demonstrated.