Double-flux variant of MixtureFluxFunctor (Abgrall & Karni 2001; Billet & Abgrall 2003; Ma, Lv & Ihme 2017): each cell's energy update uses a flux computed with its own frozen [gamma, e0] on both sides of the face, so pressure and velocity stay exactly uniform across contacts between different gases. Mass and momentum use the mean of the two sides' fluxes, which stays conservative and equals both at such contacts: both estimate the wave speeds with each side's own [gamma, e0], which makes their mass fluxes there agree for every solver, HLL and Rusanov included. face_flux holds side 0's energy flux, face_energy_1 side 1's (both as the rate of change of side 0). More...

#include <mixture_flux.h>

Public Member Functions
KOKKOS_INLINE_FUNCTION void	operator() (const uint32_t i_face) const

Public Attributes
Kokkos::View< rtype *[N_DIM]>	normals

Kokkos::View< rtype * >	face_area

Kokkos::View< int32_t *[2]>	cells_of_face

Kokkos::View< rtype * >	quad_weights

Kokkos::View< rtype ** >	face_weights

Kokkos::View< rtype **[2][N_CONSERVATIVE]>	face_solution

BoundaryData	boundaries

Kokkos::View< rtype *[N_CONSERVATIVE]>	W_cells

Kokkos::View< rtype *[2]>	frozen

Kokkos::View< rtype **[2][2]>	face_thermo

Kokkos::View< rtype **, Kokkos::LayoutStride >	cell_thermo

Kokkos::View< rtype *[N_CONSERVATIVE]>	face_flux

Kokkos::View< rtype * >	face_energy_1

Kokkos::View< rtype ** >	face_mdot

rtype	low_mach_cutoff

Kokkos::View< rtype * >	upwind

Kokkos::View< rtype *** >	quad_normals

Detailed Description

template<typename T_riemann_solver>
struct MixtureDoubleFluxFunctor< T_riemann_solver >

Double-flux variant of MixtureFluxFunctor (Abgrall & Karni 2001; Billet & Abgrall 2003; Ma, Lv & Ihme 2017): each cell's energy update uses a flux computed with its own frozen [gamma, e0] on both sides of the face, so pressure and velocity stay exactly uniform across contacts between different gases. Mass and momentum use the mean of the two sides' fluxes, which stays conservative and equals both at such contacts: both estimate the wave speeds with each side's own [gamma, e0], which makes their mass fluxes there agree for every solver, HLL and Rusanov included. face_flux holds side 0's energy flux, face_energy_1 side 1's (both as the rate of change of side 0).

Public Member Functions

Public Attributes

Detailed Description

Member Function Documentation

◆ operator()()

Member Data Documentation

◆ boundaries

◆ cell_thermo

◆ cells_of_face

◆ face_area

◆ face_energy_1

◆ face_flux

◆ face_mdot

◆ face_solution

◆ face_thermo

◆ face_weights

◆ frozen

◆ low_mach_cutoff

◆ normals

◆ quad_normals

◆ quad_weights

◆ upwind

◆ W_cells