mod_extrapolation_computer Module Reference

Computer for evaluating Extrapolation Schemes. More...

Functions/Subroutines
double precision function	ext_compute_from_values (scheme, ext_step, steps, values)
	Compute the extrapolation via the scheme by giving the associated field values.
double precision function	ext_compute_from_scheme (scheme, values)
	Compute the extrapolation of the scheme by giving the associated field values.
double precision function	ext_compute_from_field (scheme, ext_step, x_step_array, y_step_array, z_step_array, field, i, j, k, axis)
	Compute the extrapolation via the scheme by giving the associated field, the step arrays for each dimension, the indices (position) and the direction (x,y or z).
double precision function	ext_compute_from_face_field (scheme, ext_step, x_step_array, y_step_array, z_step_array, field, i, j, k, axis)
	Compute the extrapolation via the scheme by giving the associated face field, the step arrays for each dimension, the indices (position) and the direction (x,y or z).
double precision function	ext_compute_with_neumann (scheme, ext_step, steps, values, derivative_value)
	Compute the extrapolation of the scheme by giving the associated field values and the derivative value (ie. Neumann constraint).
double precision function	new_ext_compute_from_field (scheme, ext_step, x_step_array, y_step_array, z_step_array, field, i, j, k, axis, direction)
	Compute the extrapolation via the scheme by giving the associated field, the step arrays for each dimension, the indices (position) and the direction (x,y or z).
double precision function	ext_rec_compute_from_field (scheme, ext_step, x_cv_array, y_cv_array, z_cv_array, field, i, j, k, axis, direction)
	Compute the extrapolation with reconstruction via the scheme by giving the associated field, the step arrays for each dimension, the indices (position) and the direction (x,y or z).

Detailed Description

Computer for evaluating Extrapolation Schemes.

All the methods defined here are used to effectively compute/evaluate extrapolations. They take as argument a scheme that is built through the various schemes of the module (see mod_extrapolation_scheme_o1 for example).

Function/Subroutine Documentation

ext_compute_from_face_field()

double precision function mod_extrapolation_computer::ext_compute_from_face_field	(	class(t_ext_scheme), intent(inout)	scheme,
		double precision, intent(in)	ext_step,
		double precision, dimension(:), intent(in)	x_step_array,
		double precision, dimension(:), intent(in)	y_step_array,
		double precision, dimension(:), intent(in)	z_step_array,
		type(t_face_field), intent(in)	field,
		integer, intent(in)	i,
		integer, intent(in)	j,
		integer, intent(in)	k,
		integer, intent(in)	axis )

Compute the extrapolation via the scheme by giving the associated face field, the step arrays for each dimension, the indices (position) and the direction (x,y or z).

Parameters

[in]	scheme	the extrapolation scheme
[in]	ext_step	the extrapolation step (gap)
[in]	x_step_array	the discretization step array in the x direction ; x_step_array(i) is the distance from face(i+1) to face(i) (different from the cell field definition)
[in]	y_step_array	the discretization step array in the y direction
[in]	z_step_array	the discretization step array in the z direction
[in]	field	the field \(\phi\) to which to apply the scheme at the given indices.
[in]	i,j,k	the indices where to compute the FD
[in]	axis	the axis (direction) \(axis={1,2,3}\) for \({x,y,z}\)

Returns

the value of the scheme evaluated with the input values

Warning

This function is not the most efficient, especialy when used in a i,j,k loop, because it switches for each direction!

ext_compute_from_field()

double precision function mod_extrapolation_computer::ext_compute_from_field	(	class(t_ext_scheme), intent(inout)	scheme,
		double precision, intent(in)	ext_step,
		double precision, dimension(:), intent(in)	x_step_array,
		double precision, dimension(:), intent(in)	y_step_array,
		double precision, dimension(:), intent(in)	z_step_array,
		double precision, dimension(:,:,:), intent(in)	field,
		integer, intent(in)	i,
		integer, intent(in)	j,
		integer, intent(in)	k,
		integer, intent(in)	axis )

Compute the extrapolation via the scheme by giving the associated field, the step arrays for each dimension, the indices (position) and the direction (x,y or z).

Parameters

[in]	scheme	the extrapolation scheme
[in]	ext_step	the extrapolation step (gap)
[in]	x_step_array	the discretization step array in the x direction ; x_step_array(i) is the distance from cell(i) to cell(i-1)
[in]	y_step_array	the discretization step array in the y direction
[in]	z_step_array	the discretization step array in the z direction
[in]	field	the field \(\phi\) to which to apply the scheme at the given indices.
[in]	i,j,k	the indices where to compute the FD
[in]	axis	the axis (direction) \(axis={1,2,3}\) for \({x,y,z}\)

Returns

the value of the scheme evaluated with the input values

Warning

This function is not the most efficient, especialy when used in a i,j,k loop, because it switches for each direction!

ext_compute_from_scheme()

double precision function mod_extrapolation_computer::ext_compute_from_scheme	(	class(t_ext_scheme), intent(inout)	scheme,
		double precision, dimension(:), intent(in)	values )

Compute the extrapolation of the scheme by giving the associated field values.

Parameters

[in]	scheme	the extrapolation scheme (previously defined)
[in]	values	the field values to which to apply the scheme. They must be of the same dimension as the scheme

Returns

the value of the scheme evaluated with the input values

Precondition

the scheme has already been init and filled!

ext_compute_from_values()

double precision function mod_extrapolation_computer::ext_compute_from_values	(	class(t_ext_scheme), intent(inout)	scheme,
		double precision, intent(in)	ext_step,
		double precision, dimension(:), intent(in)	steps,
		double precision, dimension(:), intent(in)	values )

Compute the extrapolation via the scheme by giving the associated field values.

Parameters

[in]	scheme	the extrapolation scheme
[in]	ext_step	the extrapolation step (gap)
[in]	steps	the discretization steps, with size(steps)==size(values)-1
[in]	values	the field values to which to apply the scheme. They must be of the same dimension as the scheme

Returns

the value of the scheme evaluated with the input values

Precondition

the scheme has already been init

ext_compute_with_neumann()

double precision function mod_extrapolation_computer::ext_compute_with_neumann	(	class(t_ext_neumann_scheme), intent(inout)	scheme,
		double precision, intent(in)	ext_step,
		double precision, dimension(:), intent(in)	steps,
		double precision, dimension(:), intent(in)	values,
		double precision, intent(in)	derivative_value )

Compute the extrapolation of the scheme by giving the associated field values and the derivative value (ie. Neumann constraint).

Warning

Of course, only first derivative backward/forward schemes can be used! See

See also

Extrapolation documentation for the method's principle.

Parameters

[in]	scheme	the extrapolation scheme
[in]	ext_step	the extrapolation step (gap)
[in]	steps	the discretization steps excluding the extrapolation step included, hence with size(steps)==size(values)-1
[in]	values	the known function values to which to apply the scheme
[in]	derivative_value	the value of the derivative at the extrapolated point

Returns

the extrapolated value

Precondition

the scheme has already been init

the scheme has to be forward or backward, starting or ending exactly at index==0

ext_rec_compute_from_field()

double precision function mod_extrapolation_computer::ext_rec_compute_from_field	(	class(t_rec_scheme), intent(inout)	scheme,
		double precision, intent(in)	ext_step,
		double precision, dimension(:), intent(in)	x_cv_array,
		double precision, dimension(:), intent(in)	y_cv_array,
		double precision, dimension(:), intent(in)	z_cv_array,
		double precision, dimension(:,:,:), intent(in)	field,
		integer, intent(in)	i,
		integer, intent(in)	j,
		integer, intent(in)	k,
		integer, intent(in)	axis,
		integer, intent(in)	direction )

Compute the extrapolation with reconstruction via the scheme by giving the associated field, the step arrays for each dimension, the indices (position) and the direction (x,y or z).

Parameters

[in]	scheme	the extrapolation scheme
[in]	ext_step	the extrapolation step (gap) between the last/first node (depending on the direction)
[in]	x_cv_array	the control volumes array in the x direction
[in]	y_cv_array	the control volumesp array in the y direction
[in]	z_cv_array	the control volumes array in the z direction
[in]	field	the field \(\phi\) to which to apply the scheme at the given indices.
[in]	i,j,k	the base indices of the stencil. This has to be the last (direction=-1) or first (direction=+1) node.
[in]	axis	the axis (direction) \(axis={1,2,3}\) for \({x,y,z}\)
[in]	direction	the direction of extrapolation \(direction={-1,+1}\) for backward and forward

Returns

the value of the scheme evaluated with the input values

Warning

This function is not the most efficient, especialy when used in a i,j,k loop, because it switches for each axis!

Example

If you want to extrapolate with a 2nd order scheme, starting from node (i,j,k), at a distance of dx towards the left, you should use: ext_rec_compute_from_field( rec...o2_instance, -dx, dx_u_array, dy_v_array, dz_w_array, field, i,j,k, 1, -1 ) \par Example\ilinebr If you want to extrapolate with a 2nd order scheme, starting from node (i,j,k), at a distance of 3*dy towards the top, you should use: ext_rec_compute_from_field( rec...o2_instance, +3*dx, dx_u_array, dy_v_array, dz_w_array, field, i,j,k, 2, +1 )

new_ext_compute_from_field()

double precision function mod_extrapolation_computer::new_ext_compute_from_field	(	class(t_int_scheme), intent(inout)	scheme,
		double precision, intent(in)	ext_step,
		double precision, dimension(:), intent(in)	x_step_array,
		double precision, dimension(:), intent(in)	y_step_array,
		double precision, dimension(:), intent(in)	z_step_array,
		double precision, dimension(:,:,:), intent(in)	field,
		integer, intent(in)	i,
		integer, intent(in)	j,
		integer, intent(in)	k,
		integer, intent(in)	axis,
		integer, intent(in)	direction )

Compute the extrapolation via the scheme by giving the associated field, the step arrays for each dimension, the indices (position) and the direction (x,y or z).

Parameters

[in]	scheme	the extrapolation scheme
[in]	ext_step	the extrapolation step (gap) between the last/first node (depending on the direction)
[in]	x_step_array	the discretization step array in the x direction ; x_step_array(i) is the distance from cell(i) to cell(i-1)
[in]	y_step_array	the discretization step array in the y direction
[in]	z_step_array	the discretization step array in the z direction
[in]	field	the field \(\phi\) to which to apply the scheme at the given indices.
[in]	i,j,k	the base indices of the stencil. This has to be the last (direction=-1) or first (direction=+1) node.
[in]	axis	the axis \(axis={1,2,3}\) for \({x,y,z}\)
[in]	direction	the direction of extrapolation \(direction={-1,+1}\) for backward and forward

Returns

the value of the scheme evaluated with the input values

Note

x_step_array(i+1) is the distance between node i and node i+1. This corresponds to dx_u for cells

Warning

This function is not the most efficient, especialy when used in a i,j,k loop, because it switches for each axis!

Example

If you want to extrapolate with a 2nd order scheme, starting from node (i,j,k), at a distance of dx towards the left, you should use: new_ext_compute_from_field( int...o2_instance, -dx, dx_array, dy_array, dz_array, field, i,j,k, 1, -1 ) \par Example\ilinebr If you want to extrapolate with a 2nd order scheme, starting from node (i,j,k), at a distance of 3*dy towards the top, you should use: new_ext_compute_from_field( int...o2_instance, +3*dx, dx_array, dy_array, dz_array, field, i,j,k, 2, +1 )