Numerical simulation of a Richtmyer-Meshkov instability with an adaptive central-upwind sixth-order WENO scheme

V. K. Tritschler, X. Y. Hu, S. Hickel, N. A. Adams

Research output: Contribution to journalArticleScientificpeer-review

16 Citations (Scopus)


Two-dimensional simulations of the single-mode Richtmyer-Meshkov instability (RMI) are conducted and compared to experimental results of Jacobs and Krivets (2005 Phys. Fluids 17 034105). The employed adaptive central-upwind sixth-order weighted essentially non-oscillatory (WENO) scheme (Hu et al 2010 J. Comput. Phys. 229 8952-65) introduces only very small numerical dissipation while preserving the good shock-capturing properties of other standard WENO schemes. Hence, it is well suited for simulations with both small-scale features and strong gradients. A generalized Roe average is proposed to make the multicomponent model of Shyue (1998 J. Comput. Phys. 142 208-42) suitable for high-order accurate reconstruction schemes. A first sequence of single-fluid simulations is conducted and compared to the experiment. We find that the WENO-CU6 method better resolves small-scale structures, leading to earlier symmetry breaking and increased mixing. The first simulation, however, fails to correctly predict the global characteristic structures of the RMI. This is due to a mismatch of the post-shock parameters in single-fluid simulations when the pre-shock states are matched with the experiment. When the post-shock parameters are matched, much better agreement with the experimental data is achieved. In a sequence of multifluid simulations, the uncertainty in the density gradient associated with transition between the fluids is assessed. Thereby the multifluid simulations show a considerable improvement over the single-fluid simulations.

Original languageEnglish
Article number014016
JournalPhysica Scripta
Issue numberT155
Publication statusPublished - 1 Jul 2013
Externally publishedYes


Dive into the research topics of 'Numerical simulation of a Richtmyer-Meshkov instability with an adaptive central-upwind sixth-order WENO scheme'. Together they form a unique fingerprint.

Cite this