fluxcompute.c File Reference

Per-point calculations for getting flux. More...

#include "decs.h"

Include dependency graph for fluxcompute.c:

Go to the source code of this file.

Macros
#define	HLLCOMPUTE(cmin, cmax, U_l, U_r, F_l, F_r)   (( (cmax * F_l + cmin * F_r) - (FLUXDISSIPATION)*(cmax * cmin) * (U_r - U_l) ) / (cmax + cmin) )
	Harten-Lax-van Leer. More...
#define	LAXFCOMPUTE(ctop, U_l, U_r, F_l, F_r)   (0.5 * ( (F_l + F_r) - (FLUXDISSIPATION)*ctop * (U_r - U_l) ) )
	Lax-Friedrichs with Rusanov wave speed. More...
#define	LFCOMPUTE(crus, U_l, U_r, F_l, F_r)   (0.5*( (F_l+F_r) - (FLUXDISSIPATION)*crus*(U_r - U_l)))
	Lax-Friedrichs flux. More...
#define	UHALF(crus, U_l, U_r, F_l, F_r)   (0.5*((U_l+U_r)-(FLUXDISSIPATION)*(F_r-F_l)/crus))
	mid-step for 2-step LW-flux More...
#define	LWCOMPUTE(crus, U_l, U_r, F_l, Fhalf, F_r)   (Fhalf)
	2-step Lax-Wendroff flux More...
#define	FORCECOMPUTE(crus, U_l, U_r, F_l, Fhalf, F_r)   (0.25*( (F_l+F_r) + 2.0*Fhalf - (FLUXDISSIPATION)*crus*(U_r - U_l)))
	FORCE. More...
#define	GFORCECOMPUTE(cg, crus, U_l, U_r, F_l, Fhalf, F_r)   ( (LWCOMPUTE(crus,U_l,U_r,F_l,Fhalf,F_r) + cg*LFCOMPUTE(crus,U_l,U_r,F_l,F_r))/(1.0+cg) )
	GFORCE // Toro & Titarev JCP 2006. More...
#define	HLLLAXF1(cmin, cmax, ctop, f_s, U_l, U_r, F_l, F_r)   ( HLLCOMPUTE(cmin,cmax,U_l,U_r,F_l,F_r)*f_s + LAXFCOMPUTE(ctop,U_l,U_r,F_l,F_r)*(1.0-f_s) )
	use more HLL as f_s from 0 to 1, use more LAXF as f_s from 1 to 0 More...
#define	MINDPOP   (0.0)
#define	MAXDPOP   (0.3)
#define	USE_CORRECTED_STATES   0
	e.g. double rarefaction Einfelt test (test2 in Liska & Wendroff 2003) leads to F at interface corresponding to negative pressure. This is supposed to correct for that. Leads to slightly better solution except for near center More...
#define	MUSTACOEF   (1.0)
#define	NUMMUSTAITERS   1
	number of single-cell musta iterations More...
#define	DOMULTICELL   0
	whether to do mutli-cell version of MUSTA More...
#define	NUMMULTIMUSTAITERS   2
	number of multicell musta iterations More...
#define	NUMLOCALCELLS   2
	NUMMULTIMUSTAITERS=2 and NUMLOCALCELLS=1 is simplest scheme multicell much much slower than single cell version number of musta cells. More...
#define	MULTIMUSTACOEF   (0.9)
#define	MUSTAFORCE   0
#define	MUSTALAXF   1
#define	MUSTAHLL   2
#define	WHICHFLUX   MUSTAHLL
#define	HLLBOUNDMUSTA   0
#define	MUSTAVERSION   0

Functions
int	flux_compute_general (int i, int j, int k, int dir, struct of_geom *ptrgeom, FTYPE CUf, FTYPE *p_c, FTYPE *p_l, FTYPE *p_r, FTYPE *F, FTYPE *ctopallptr)
	actually compute the flux and ctop for dt calculation More...
int	p2SFUevolve (int dir, int isleftright, FTYPE *p, struct of_geom *ptrgeom, struct of_state **ptrstate, FTYPE *F, FTYPE *U)
	P->sate,flux,U. More...
int	p2SFUevolveall (int dir, int isleftright, FTYPE *p, struct of_geom *ptrgeom, struct of_state **ptrstate, FTYPE *FPAKE, FTYPE *FEN, FTYPE *FEM, FTYPE *FRAD, FTYPE *UPAKE, FTYPE *UEN, FTYPE *UEM, FTYPE *URAD)
int	flux_compute (int i, int j, int k, int dir, struct of_geom *geom, FTYPE *cminmax_l, FTYPE *cminmax_r, FTYPE *cminmax, FTYPE ctopmhd, FTYPE *cminmaxrad_l, FTYPE *cminmaxrad_r, FTYPE *cminmaxrad, FTYPE ctoprad, FTYPE CUf, FTYPE *p_l, FTYPE *p_r, FTYPE *U_l, FTYPE *U_r, FTYPE *F_l, FTYPE *F_r, FTYPE *F)
	actually compute flux from given data GODMARK: right now if splitmaem==1 assumes F is linear in FMA and FEM. More...

Detailed Description

Per-point calculations for getting flux.

ideas: GODMARK: Consider ZIP-average for non-diffusive part of LAXF or HLL flux Chacon (2004) Computer Physics Communications

Definition in file fluxcompute.c.

Macro Definition Documentation

#define DOMULTICELL   0

whether to do mutli-cell version of MUSTA

Definition at line 711 of file fluxcompute.c.

#define FORCECOMPUTE	(		crus,
			U_l,
			U_r,
			F_l,
			Fhalf,
			F_r
	)		(0.25*( (F_l+F_r) + 2.0*Fhalf - (FLUXDISSIPATION)*crus*(U_r - U_l)))

FORCE.

Definition at line 331 of file fluxcompute.c.

#define GFORCECOMPUTE	(		cg,
			crus,
			U_l,
			U_r,
			F_l,
			Fhalf,
			F_r
	)		( (LWCOMPUTE(crus,U_l,U_r,F_l,Fhalf,F_r) + cg*LFCOMPUTE(crus,U_l,U_r,F_l,F_r))/(1.0+cg) )

GFORCE // Toro & Titarev JCP 2006.

Definition at line 334 of file fluxcompute.c.

#define HLLBOUNDMUSTA   0

Definition at line 734 of file fluxcompute.c.

#define HLLCOMPUTE	(		cmin,
			cmax,
			U_l,
			U_r,
			F_l,
			F_r
	)		(( (cmax * F_l + cmin * F_r) - (FLUXDISSIPATION)*(cmax * cmin) * (U_r - U_l) ) / (cmax + cmin) )

Harten-Lax-van Leer.

Definition at line 313 of file fluxcompute.c.

#define HLLLAXF1	(		cmin,
			cmax,
			ctop,
			f_s,
			U_l,
			U_r,
			F_l,
			F_r
	)		( HLLCOMPUTE(cmin,cmax,U_l,U_r,F_l,F_r)*f_s + LAXFCOMPUTE(ctop,U_l,U_r,F_l,F_r)*(1.0-f_s) )

use more HLL as f_s from 0 to 1, use more LAXF as f_s from 1 to 0

Definition at line 338 of file fluxcompute.c.

#define LAXFCOMPUTE	(		ctop,
			U_l,
			U_r,
			F_l,
			F_r
	)		(0.5 * ( (F_l + F_r) - (FLUXDISSIPATION)*ctop * (U_r - U_l) ) )

Lax-Friedrichs with Rusanov wave speed.

Definition at line 316 of file fluxcompute.c.

#define LFCOMPUTE	(		crus,
			U_l,
			U_r,
			F_l,
			F_r
	)		(0.5*( (F_l+F_r) - (FLUXDISSIPATION)*crus*(U_r - U_l)))

Lax-Friedrichs flux.

Definition at line 322 of file fluxcompute.c.

#define LWCOMPUTE	(		crus,
			U_l,
			U_r,
			F_l,
			Fhalf,
			F_r
	)		(Fhalf)

2-step Lax-Wendroff flux

Definition at line 328 of file fluxcompute.c.

#define MAXDPOP   (0.3)

#define MINDPOP   (0.0)

#define MULTIMUSTACOEF   (0.9)

Definition at line 718 of file fluxcompute.c.

#define MUSTACOEF   (1.0)

Definition at line 594 of file fluxcompute.c.

#define MUSTAFORCE   0

Definition at line 721 of file fluxcompute.c.

#define MUSTAHLL   2

Definition at line 723 of file fluxcompute.c.

#define MUSTALAXF   1

Definition at line 722 of file fluxcompute.c.

#define MUSTAVERSION   0

#define NUMLOCALCELLS   2

NUMMULTIMUSTAITERS=2 and NUMLOCALCELLS=1 is simplest scheme multicell much much slower than single cell version number of musta cells.

Definition at line 717 of file fluxcompute.c.

#define NUMMULTIMUSTAITERS   2

number of multicell musta iterations

Definition at line 713 of file fluxcompute.c.

#define NUMMUSTAITERS   1

number of single-cell musta iterations

Definition at line 708 of file fluxcompute.c.

#define UHALF	(		crus,
			U_l,
			U_r,
			F_l,
			F_r
	)		(0.5*((U_l+U_r)-(FLUXDISSIPATION)*(F_r-F_l)/crus))

mid-step for 2-step LW-flux

Definition at line 325 of file fluxcompute.c.

#define USE_CORRECTED_STATES   0

e.g. double rarefaction Einfelt test (test2 in Liska & Wendroff 2003) leads to F at interface corresponding to negative pressure. This is supposed to correct for that. Leads to slightly better solution except for near center

Definition at line 544 of file fluxcompute.c.

#define WHICHFLUX   MUSTAHLL

Definition at line 725 of file fluxcompute.c.

Function Documentation

int flux_compute	(	int	i,
		int	j,
		int	k,
		int	dir,
		struct of_geom *	geom,
		FTYPE *	cminmax_l,
		FTYPE *	cminmax_r,
		FTYPE *	cminmax,
		FTYPE	ctopmhd,
		FTYPE *	cminmaxrad_l,
		FTYPE *	cminmaxrad_r,
		FTYPE *	cminmaxrad,
		FTYPE	ctoprad,
		FTYPE	CUf,
		FTYPE *	p_l,
		FTYPE *	p_r,
		FTYPE *	U_l,
		FTYPE *	U_r,
		FTYPE *	F_l,
		FTYPE *	F_r,
		FTYPE *	F
	)

actually compute flux from given data GODMARK: right now if splitmaem==1 assumes F is linear in FMA and FEM.

In general Riemann solver should return FMA and FEM in parts

Definition at line 346 of file fluxcompute.c.

Here is the call graph for this function:

Here is the caller graph for this function:

int flux_compute_general	(	int	i,
		int	j,
		int	k,
		int	dir,
		struct of_geom *	ptrgeom,
		FTYPE	CUf,
		FTYPE *	p_c,
		FTYPE *	p_l,
		FTYPE *	p_r,
		FTYPE *	F,
		FTYPE *	ctopallptr
	)

actually compute the flux and ctop for dt calculation

actually compute flux from given data

Definition at line 25 of file fluxcompute.c.

Here is the call graph for this function:

Here is the caller graph for this function:

int p2SFUevolve	(	int	dir,
		int	isleftright,
		FTYPE *	p,
		struct of_geom *	ptrgeom,
		struct of_state **	ptrstate,
		FTYPE *	F,
		FTYPE *	U
	)

P->sate,flux,U.

Definition at line 199 of file fluxcompute.c.

Here is the call graph for this function:

Here is the caller graph for this function:

int p2SFUevolveall	(	int	dir,
		int	isleftright,
		FTYPE *	p,
		struct of_geom *	ptrgeom,
		struct of_state **	ptrstate,
		FTYPE *	FPAKE,
		FTYPE *	FEN,
		FTYPE *	FEM,
		FTYPE *	FRAD,
		FTYPE *	UPAKE,
		FTYPE *	UEN,
		FTYPE *	UEM,
		FTYPE *	URAD
	)

Definition at line 218 of file fluxcompute.c.

Here is the call graph for this function:

Here is the caller graph for this function: