Modélisation et simulation d'écoulements compressibles et diphasiques
Analyse numérique
Ondes de choc
LBM
Publications scientifiques au M2P2
2025
Ksenia Kozhanova, Yannick Hoarau, Eric Goncalves da Silva. A 3D numerical strategy for the computations of shock-induced bubble collapse near a wall. Computers and Fluids, 2025, 293, pp.106609. ⟨10.1016/j.compfluid.2025.106609⟩. ⟨hal-05017998⟩ Plus de détails...
The importance of modelling two-phase flows involving shock waves arises from many engineering and medical applications. The presence of strong shock waves and their interactions with bubble interfaces, the high density ratio between phases and the large variation of material properties makes the resolution of such problems a complicated task for the numerical methods. While the variety of numerical techniques to solve these problems exist, e.g. the sharp interface or the diffuse interface methods, these strategies can lead to spurious oscillations of the solution near the interface. It is well known that it is difficult to achieve both a high order accuracy of the scheme and the monotonicity of the solution. In this paper a four-equation two-phase model is employed and integrated in an explicit fully parallelised finite-volume solver with HLLC numerical scheme coupled with WENO reconstruction methods and Hancock predictor–corrector scheme and non-uniform mesh based on stretching function in order to compute a 3D shock-induced bubble collapse near a wall. The novelty of our work is improved accuracy of computations of such a problem with optimised computational cost thanks to the non-uniform mesh introduction in 3D computations.
Ksenia Kozhanova, Yannick Hoarau, Eric Goncalves da Silva. A 3D numerical strategy for the computations of shock-induced bubble collapse near a wall. Computers and Fluids, 2025, 293, pp.106609. ⟨10.1016/j.compfluid.2025.106609⟩. ⟨hal-05017998⟩
Ksenia Kozhanova, Song Zhao, Raphaël Loubère, Pierre Boivin. A hybrid a posteriori MOOD limited lattice Boltzmann method to solve compressible fluid flows – LBMOOD. Journal of Computational Physics, 2025, 521, Part 2, pp.113570. ⟨10.1016/j.jcp.2024.113570⟩. ⟨hal-04802259⟩ Plus de détails...
In this paper we blend two lattice-Boltzmann (LB) numerical schemes with an a posteriori Multi-dimensional Optimal Order Detection (MOOD) paradigm to solve hyperbolic systems of conservation laws in 1D and 2D. The first LB scheme is robust to the presence of shock waves but lacks accuracy on smooth flows. The second one has a second-order of accuracy but develops non-physical oscillations when solving steep gradients. The MOOD paradigm produces a hybrid LB scheme via smooth and positivity detectors allowing to gather the best properties of the two LB methods within one scheme. Indeed, the resulting scheme presents second order of accuracy on smooth solutions, essentially non-oscillatory behaviour on irregular ones, and, an ‘almost fail-safe’ property concerning positivity issues. The numerical results on a set of sanity test cases and demanding ones are presented assessing the appropriate behaviour of the hybrid LBMOOD scheme in 1D and 2D.
Ksenia Kozhanova, Song Zhao, Raphaël Loubère, Pierre Boivin. A hybrid a posteriori MOOD limited lattice Boltzmann method to solve compressible fluid flows – LBMOOD. Journal of Computational Physics, 2025, 521, Part 2, pp.113570. ⟨10.1016/j.jcp.2024.113570⟩. ⟨hal-04802259⟩
