A1 Refereed original research article in a scientific journal
Physically consistent formulations of split convective terms for turbulent compressible multi-component flows
Authors: Wang, Ye; Wehrfritz, Armin; Hawkes, Evatt R.
Publisher: Elsevier BV
Publication year: 2025
Journal: Journal of Computational Physics
Journal name in source: Journal of Computational Physics
Article number: 114269
Volume: 540
ISSN: 0021-9991
eISSN: 1090-2716
DOI: https://doi.org/10.1016/j.jcp.2025.114269
Web address : https://doi.org/10.1016/j.jcp.2025.114269
Self-archived copy’s web address: https://research.utu.fi/converis/portal/detail/Publication/499812028
We analyse the properties and characteristics of kinetic-energy-preserving, entropy-preserving, and pressure-equilibrium-preserving split convective forms for compressible multi-component flows. The results show that such schemes offer improved pressure-equilibrium-preserving properties and numerical stability compared to most other existing schemes, but also that the preservation of pressure equilibrium is not guaranteed for flows with varying specific heats. Furthermore, for the convective terms in species mass fraction transport equations, some split forms may fail to preserve key physical properties discretely. We construct a formulation for the species convective terms that consistently maintains these key physical properties, including species mass conservation, uniform mass fraction preservation, and temperature-equilibrium preservation. The capability of the proposed scheme in maintaining these properties is demonstrated analytically and tested in one-dimensional advection problems. Last, the proposed scheme is compared with schemes that do not satisfy these properties in under-resolved simulations of a modified inviscid Taylor–Green vortex flow. The results show improved performance of the proposed scheme and highlight the importance of a convective scheme for the species mass fractions to be able to consistently preserve these physical properties in a discrete sense.
Downloadable publication This is an electronic reprint of the original article. |  |
Funding information in the publication:
This work was supported by the Australian Government through the Australian Research Council’s Discovery Projects funding scheme (project DP200103535). A pool of computational resources was provided by the Australian Government through the Pawsey Supercomputing Centre and the National Computational Infrastructure under the National Computational Merit Allocation Scheme, and by the University of New South Wales.
