coloredblusgs.c File Reference

Colored Block LU-SGS time integration method for unstructured grid. More...

#include <stdio.h>
#include <omp.h>
#include "coloredblusgs.h"
#include "flux.h"
#include "inverse.h"

Include dependency graph for coloredblusgs.c:

Go to the source code of this file.

Functions
void	ns_parallel_pre_blusgs (int neles, int nfvars, int nface, double factor, double *fnorm_vol, double *dt, double *diag, double *fjmat)
void	rans_parallel_pre_blusgs (int neles, int nvars, int nfvars, int nface, double factor, double betast, double *fnorm_vol, double *uptsb, double *dt, double *tdiag, double *tjmat, double *dsrc)
void	ns_parallel_block_sweep (int n0, int ne, int neles, int nfvars, int nface, int *nei_ele, int *icolor, int *lcolor, double *fnorm_vol, double *rhsb, double *dub, double *diag, double *fjmat)
void	rans_parallel_block_sweep (int n0, int ne, int neles, int nvars, int nfvars, int nface, int *nei_ele, int *icolor, int *lcolor, double *fnorm_vol, double *rhsb, double *dub, double *tdiag, double *tjmat)
void	parallel_update (int neles, int nvars, double *uptsb, double *dub, double *subres)

Detailed Description

Colored Block LU-SGS time integration method for unstructured grid.

======================================================================================================================

Block LU-SGS time integration method with Multi-coloring algorithm.
Multi-thread computation is enabled using <omp.h> header file.
In contrast to LU-SGS method, Block LU-SGS method uses block operator instead of spectral radius.

Note

In case of RANS equations, add rans_parallel_pre_blusgs function right after the ns_parallel_pre_blusgs function.
There is only block_sweep function instead of block_lower_sweep or block_upper_sweep in blusgs.c.
To execute properly, n0 and ne arguments must be input in appropriate order.

Author

• Namhyoung Kim (knhks.nosp@m.e@in.nosp@m.ha.ed.nosp@m.u), Department of Aerospace Engineering, Inha University
• Jin Seok Park (jinse.nosp@m.ok.p.nosp@m.ark@i.nosp@m.nha..nosp@m.ac.kr), Department of Aerospace Engineering, Inha University

Date

Nov 2024

Version

1.0

Copyright

Copyright (c) 2024, Namhyoung Kim and Jin Seok Park, Inha University, All rights reserved.

License

This project is release under the terms of the MIT License (see LICENSE file).

---

Definition in file coloredblusgs.c.

Function Documentation

ns_parallel_block_sweep()

void ns_parallel_block_sweep	(	int	n0,
		int	ne,
		int	neles,
		int	nfvars,
		int	nface,
		int *	nei_ele,
		int *	icolor,
		int *	lcolor,
		double *	fnorm_vol,
		double *	rhsb,
		double *	dub,
		double *	diag,
		double *	fjmat
	)

Lower/Upper sweep of Block LU-SGS with multi-coloring algorithm.
This function is used for Euler or Navier-Stokes equations, which has the same flux shape.
solution array is stored in dub array.

Note

The last argument array, fjmat is NOT identical with ns_serial_pre_blusgs function.
For more details, refer to the Block LU-SGS in the document.

Definition at line 155 of file coloredblusgs.c.

Here is the call graph for this function:

ns_parallel_pre_blusgs()

void ns_parallel_pre_blusgs	(	int	neles,
		int	nfvars,
		int	nface,
		double	factor,
		double *	fnorm_vol,
		double *	dt,
		double *	diag,
		double *	fjmat
	)

This function computes diagonal matrices of the implicit operator using multi-thread computing.
In Block LU-SGS method, implicit operator is approximated with block operator. Diagonal matrices is composed of block operator matrix, which size is n-by-n. n is the number of conservative variables in Navier-Stokes equations.

Definition at line 38 of file coloredblusgs.c.

Here is the call graph for this function:

parallel_update()

void parallel_update	(	int	neles,
		int	nvars,
		double *	uptsb,
		double *	dub,
		double *	subres
	)

solution array is updated by adding \(\Delta Q\).

Definition at line 275 of file coloredblusgs.c.

Here is the caller graph for this function:

rans_parallel_block_sweep()

void rans_parallel_block_sweep	(	int	n0,
		int	ne,
		int	neles,
		int	nvars,
		int	nfvars,
		int	nface,
		int *	nei_ele,
		int *	icolor,
		int *	lcolor,
		double *	fnorm_vol,
		double *	rhsb,
		double *	dub,
		double *	tdiag,
		double *	tjmat
	)

Lower/Upper sweep of Block LU-SGS with multi-coloring algorithm.
This function is used for RANS equations. solution array is stored in dub array.

Note

The last argument array, tjmat is NOT identical with rans_parallel_pre_blusgs function.
For more details, refer to the Block LU-SGS in the document.

Definition at line 216 of file coloredblusgs.c.

Here is the call graph for this function:

rans_parallel_pre_blusgs()

void rans_parallel_pre_blusgs	(	int	neles,
		int	nvars,
		int	nfvars,
		int	nface,
		double	factor,
		double	betast,
		double *	fnorm_vol,
		double *	uptsb,
		double *	dt,
		double *	tdiag,
		double *	tjmat,
		double *	dsrc
	)

This function computes diagonal matrices of the implicit operator of RANS equations. In Block LU-SGS method, implicit operator is approximated with block operator. Diagonal matrices is composed of block operator matrix, which size is n-by-n. n is the number of turbulent variables in RANS equations.

Definition at line 88 of file coloredblusgs.c.

Here is the call graph for this function: