#include "variables.h"
#include "logging.h"

Include dependency graph for Filter.h:

This graph shows which files directly or indirectly include this file:

Functions
double	ApplyLESTestFilter (LESTestFilterKernel kernel, double values[3][3][3], double weights[3][3][3])
	Applies a numerical "test filter" to a 3x3x3 stencil of data points.

PetscErrorCode	ApplyLESTestFilterSymTensor (LESTestFilterKernel kernel, SymTensor values[3][3][3], double weights[3][3][3], SymTensor *filtered)
	Applies the test filter to all six components of a symmetric tensor.

Function Documentation

◆ ApplyLESTestFilter()

double ApplyLESTestFilter	(	LESTestFilterKernel	kernel,
		double	values[3][3][3],
		double	weights[3][3][3]
	)

Applies a numerical "test filter" to a 3x3x3 stencil of data points.

EXPLANATION of "Test Filter": In the dynamic Smagorinsky LES model the model coefficient is not prescribed. It is calculated dynamically by sampling the turbulent energy cascade at two different scales. The first scale is the grid filter itself, implicit in the discretization. The second, larger scale is this explicit "test filter". By comparing how the resolved flow behaves at the two scales, the model determines how much dissipation the unresolved motions require.

The kernel is selected by the caller rather than read from the simulation context, so a filter can be exercised on 27 numbers without a live run.

Parameters

[in]	kernel	Discrete stencil to apply.
[in]	values	The 3x3x3 array of scalar values at the stencil points, indexed `[k][j][i]` in the DMDA convention.
[in]	weights	The 3x3x3 array of cell volume weights. Ignored by LES_TEST_FILTER_SIMPSON_IK, which assumes uniform spacing in its homogeneous plane.

Returns: The filtered value, or zero when every weight vanishes because the stencil lies entirely inside a solid body.

Applies a numerical "test filter" to a 3x3x3 stencil of data points.

Full API contract (arguments, ownership, side effects) is documented with the header declaration in include/Filter.h.

See also: ApplyLESTestFilter()

Definition at line 123 of file Filter.c.

{
    // This function acts as a dispatcher over the configured stencil.
    switch (kernel) {
    case LES_TEST_FILTER_SIMPSON_IK:
        // The caller has declared the i and k directions homogeneous, so the more
        // accurate Simpson stencil applies. Volume weights are irrelevant there.
        return ApplySimpsonRuleHomogeneousFilter(values);
    case LES_TEST_FILTER_VOLUME_WEIGHTED_BOX:
    default:
        // Default for general, non-uniform, curvilinear grids. The volume-weighted
        // box filter keeps the average correct when neighbouring cells differ in size.
        return ApplyVolumeWeightedBoxFilter(values, weights);
    }
}

Here is the call graph for this function:

Here is the caller graph for this function:

◆ ApplyLESTestFilterSymTensor()

PetscErrorCode ApplyLESTestFilterSymTensor	(	LESTestFilterKernel	kernel,
		SymTensor	values[3][3][3],
		double	weights[3][3][3],
		SymTensor *	filtered
	)

Applies the test filter to all six components of a symmetric tensor.

Filtering is linear, so applying it componentwise is exact; this wrapper exists so callers state the intent once instead of repeating six near-identical calls whose only difference is easy to mistype.