Abstract

Balsara's genuinely multidimensional Riemann solvers, which possess a simple closed-form and high efficiency in the low-order cases, realize their multidimensional effect successfully by introducing the vertex-based framework. However, traditional higher-order (third-order and above) reconstruction methods built for the structured grids become ambiguous when applied directly into these solvers. Nowadays, there is little research on the higher-order reconstruction methods for such Riemann solvers besides Balsara's work by adopting the ADER-WENO schemes. Based on Balsara's work, we conduct research on the multidimensional higher-order reconstruction procedure in this study. This paper investigates two third-order reconstruction methods with the WENO approach for the genuinely two-dimensional HLLE Riemann solver. The critical idea of the reconstruction methods is to utilize the solution or its derivative to construct the spatial two-dimensional interpolation polynomials with the third-order accuracy. The first reconstruction method denoted by TWENO constructs the polynomials by the Taylor expansion, while the second one denoted by HWENO adopts the Hermite interpolation polynomials. By combining the polynomials with the WENO limiter, these schemes can provide required reconstructed values at the midpoints and corner-points of the cell interfaces for the framework and avoid oscillations near discontinuities. A discontinuity-detector technology is adopted to improve the computational efficiency of these reconstruction methods in simulating complex flows. Several numerical test cases are conducted and show that these third-order methods achieve their design accuracy and capture the shock waves with no overshoots or oscillations. The multidimensional nature of the solutions is kept well by the TWENO/HWENO methods, and the computational time is reduced by about 30% with the discontinuity-detector technology. Moreover, it is found that the genuinely two-dimensional Riemann solver fits the third-order reconstruction methods naturally due to its unique application of a Simpson rule.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call