Efficient Arbitrary-High-Order TENO Schemes with Local Adaptive Dissipation for Compressible Flow Simulation on Unstructured Meshes
摘要
For compressible flow simulations involving both shock waves and turbulence, the competing requirements render it challenging to develop high-order numerical methods capable of capturing the discontinuities sharply and resolving the turbulence with high spectral resolution. In this paper, an efficient class of high-order TENO schemes with local adaptive dissipation for compressible flow simulation on unstructured meshes is proposed based on three new concepts: (1) a novel reliable troubled-cell indicator is proposed for the unstructured finite-volume method without case-sensitive parameter to tune; (2) different from the classical shock-capturing schemes for unstructured meshes, which conduct characteristic decomposition at each cell interface, an efficient hybrid weighting strategy is proposed by recasting the high-order linear scheme based on conserved variables for smooth flow scales and invoking the nonlinear TENO weighting process in characteristic space for non-smooth flow scales; (3) noticing that the low-order undivided difference deployed in the calculation of the new indicator is more effective in terms of separating the high-wavenumber fluctuations from the genuine discontinuities than the high-order difference, a new adaptive dissipation control strategy is introduced to combine the good numerical robustness for shock waves with the low-dissipation property for broadband physical fluctuations. Without the necessity of parameter tuning case by case, a set of benchmark simulations reveals that the proposed TENO-E scheme features robust shock-capturing capability and state-of-the-art high-resolution properties for highly compressible flows involving strong shock waves and a wide range of flow scales. Moreover, the proposed scheme is substantially less computationally expensive than the straightforward deployment of classical shock-capturing schemes, and thus is promising for high-fidelity DNS/LES simulation of more complex practical engineering flows.