Face linear system preparation

Face linear system preparation of a vectorial equation defined on faces. More...

Namespaces
module	type_face_ls_map
	Type decalaration associated to the mapping of a face vector linear system on the numerical grid.

Functions
subroutine	mod_add_diagonal_face_matrix::add_diagonal_face_matrix (matrix, navier_stencil_size, work_array, navier_ls_map)
	Add a term to the diagonal of a matrix for an equation defined on faces.
subroutine	mod_add_rhs_face::add_rhs_face (rhs, work_array, navier_ls_map)
	Add arrays to Right-Hand-Side corresponding to an equation discretized on faces.
subroutine	mod_finalize_face_linear_system::finalize_face_linear_system (matrix, rhs, stencil_size, equation_face_ls_map)
	Detect unset rows and set them ready to pass ls solvers.
pure subroutine	mod_initialize_face_linear_system::initialize_face_linear_system (matrix, rhs, ls_map)
	Allocate matrix and right-hand side of a face linear system.
subroutine	mod_time_advance_explicitly_face::time_advance_explicitly_face (matrix, rhs, work_array)
	Determine \( \mathbf{u}^{n+1} \) for an equation discretized on faces and with a fully explicit implementation.

Detailed Description

Function Documentation

add_rhs_face()

subroutine mod_add_rhs_face::add_rhs_face	(	double precision, dimension(:), intent(inout)	rhs,
		type(t_face_field), intent(in)	work_array,
		type(t_face_ls_map), intent(in)	navier_ls_map )

!!

Note

• The add_rhs_face routine takes cell field elements and adds them to the right-hand side of a linear system.

finalize_face_linear_system()

subroutine mod_finalize_face_linear_system::finalize_face_linear_system	(	double precision, dimension(:), intent(inout)	matrix,
		double precision, dimension(:), intent(inout)	rhs,
		integer, intent(in)	stencil_size,
		type(t_face_ls_map), intent(in)	equation_face_ls_map )

Unset rows are detected by checking diagonal element values of that equals to 0d0. Then set the row to identity and the right-hand side to 0.

Todo

improve unused nodes values (extrapolation)

time_advance_explicitly_face()

subroutine mod_time_advance_explicitly_face::time_advance_explicitly_face	(	double precision, dimension(:), intent(in)	matrix,
		double precision, dimension(:), intent(in)	rhs,
		type(t_face_field), intent(inout)	work_array )

!!

Note

• The time_advance_explicitly_face routine solves \( \alpha_{i,j,k} \u^{n+1} = RHS_{i,j,k} \) and applies the appropriate boundary conditions.