#include <source/euler/limiter.h>

Public Member Functions
Stencil-based computation of bounds
Intended usage: LimiterView<dim, Number> limiter_view; for (unsigned int i = n_internal; i < n_owned; ++i) { // ... limiter_view.reset(pv, i, U_i, flux_i); for (unsigned int col_idx = 1; col_idx < row_length; ++col_idx) { // ... limiter_view.accumulate(pv, js, U_j, flux_j, scaled_c_ij, affine_shift); } limiter_view.bounds(hd_i); } ryujin::Euler::LimiterView Definition limiter.h:201 ryujin::Euler::LimiterView::reset DEAL_II_HOST_DEVICE void reset(const PrecomputedVectorView &pv, const unsigned int i, const state_type &U_i, const flux_contribution_type &flux_i) Definition limiter.h:473 ryujin::Euler::LimiterView::accumulate DEAL_II_HOST_DEVICE void accumulate(const PrecomputedVectorView &pv, const unsigned int js, const state_type &U_j, const flux_contribution_type &flux_j, const dealii::Tensor< 1, dim, Number > &scaled_c_ij, const state_type &affine_shift) Definition* limiter.h:499 ryujin::Euler::LimiterView::bounds DEAL_II_HOST_DEVICE Bounds bounds(const Number hd_i) const Definition limiter.h:550
DEAL_II_HOST_DEVICE void	reset (const PrecomputedVectorView &pv, const unsigned int i, const state_type &U_i, const flux_contribution_type &flux_i)

DEAL_II_HOST_DEVICE void	accumulate (const PrecomputedVectorView &pv, const unsigned int *js, const state_type &U_j, const flux_contribution_type &flux_j, const dealii::Tensor< 1, dim, Number > &scaled_c_ij, const state_type &affine_shift)

DEAL_II_HOST_DEVICE Bounds	bounds (const Number hd_i) const

Convex limiter
DEAL_II_HOST_DEVICE std::tuple< Number, bool >	limit (const Bounds &bounds, const state_type &U, const state_type &P, const Number t_min=Number(0.), const Number t_max=Number(1.)) const

Typedefs and constexpr constants
using	View = HyperbolicSystemView< dim, Number, MemorySpace >

using	ScalarNumber = typename View::ScalarNumber

using	state_type = typename View::state_type

using	flux_contribution_type = typename View::flux_contribution_type

using	precomputed_type = typename View::precomputed_type

using	PrecomputedVectorView = typename View::PrecomputedVectorView

static constexpr auto	problem_dimension = View::problem_dimension

Computation and manipulation of bounds
using	Bounds = std::array< Number, n_bounds >

static constexpr unsigned int	n_bounds = 3

	LimiterView (const View &view, const Limiter< ScalarNumber > &limiter)

DEAL_II_HOST_DEVICE_ALWAYS_INLINE unsigned int	iterations () const

DEAL_II_HOST_DEVICE_ALWAYS_INLINE ScalarNumber	newton_tolerance () const

DEAL_II_HOST_DEVICE_ALWAYS_INLINE unsigned int	newton_max_iterations () const

DEAL_II_HOST_DEVICE_ALWAYS_INLINE ScalarNumber	relaxation_factor () const

DEAL_II_HOST_DEVICE Bounds	projection_bounds_from_state (const PrecomputedVectorView &pv, const unsigned int i, const state_type &U_i) const

DEAL_II_HOST_DEVICE Bounds	combine_bounds (const Bounds &bounds_left, const Bounds &bounds_right) const

DEAL_II_HOST_DEVICE Bounds	fully_relax_bounds (const Bounds &bounds, const Number &hd) const

Detailed Description

template<int dim, typename Number, typename MemorySpace>
class ryujin::Euler::LimiterView< dim, Number, MemorySpace >

A view of the Limiter that makes the interface available for a given dimension dim and choice of number type Number (which can be a scalar float, or double, as well as a VectorizedArray holding packed scalars).

Definition at line 200 of file limiter.h.

Member Typedef Documentation

◆ View

template<int dim, typename Number , typename MemorySpace >

using ryujin::Euler::LimiterView< dim, Number, MemorySpace >::View = HyperbolicSystemView<dim, Number, MemorySpace>

Definition at line 213 of file limiter.h.

◆ ScalarNumber

template<int dim, typename Number , typename MemorySpace >

using ryujin::Euler::LimiterView< dim, Number, MemorySpace >::ScalarNumber = typename View::ScalarNumber

Definition at line 215 of file limiter.h.

◆ state_type

template<int dim, typename Number , typename MemorySpace >

using ryujin::Euler::LimiterView< dim, Number, MemorySpace >::state_type = typename View::state_type

Definition at line 219 of file limiter.h.

◆ flux_contribution_type

template<int dim, typename Number , typename MemorySpace >

using ryujin::Euler::LimiterView< dim, Number, MemorySpace >::flux_contribution_type = typename View::flux_contribution_type

Definition at line 221 of file limiter.h.

◆ precomputed_type

template<int dim, typename Number , typename MemorySpace >

using ryujin::Euler::LimiterView< dim, Number, MemorySpace >::precomputed_type = typename View::precomputed_type

Definition at line 223 of file limiter.h.

◆ PrecomputedVectorView

template<int dim, typename Number , typename MemorySpace >

using ryujin::Euler::LimiterView< dim, Number, MemorySpace >::PrecomputedVectorView = typename View::PrecomputedVectorView

Definition at line 225 of file limiter.h.

◆ Bounds

template<int dim, typename Number , typename MemorySpace >

using ryujin::Euler::LimiterView< dim, Number, MemorySpace >::Bounds = std::array<Number, n_bounds>

Array type used to store accumulated bounds.

Definition at line 240 of file limiter.h.

Constructor & Destructor Documentation

◆ LimiterView()

template<int dim, typename Number , typename MemorySpace >

ryujin::Euler::LimiterView< dim, Number, MemorySpace >::LimiterView	(	const View &	view,
		const Limiter< ScalarNumber > &	limiter
	)

inline

Constructor taking a HyperbolicSystemView and a Limiter object as arguments

Definition at line 246 of file limiter.h.

Member Function Documentation

◆ iterations()

template<int dim, typename Number , typename MemorySpace >

DEAL_II_HOST_DEVICE_ALWAYS_INLINE unsigned int ryujin::Euler::LimiterView< dim, Number, MemorySpace >::iterations ( ) const

inline

Return the number of limiter iterations.

Definition at line 255 of file limiter.h.

◆ newton_tolerance()

template<int dim, typename Number , typename MemorySpace >

DEAL_II_HOST_DEVICE_ALWAYS_INLINE ScalarNumber ryujin::Euler::LimiterView< dim, Number, MemorySpace >::newton_tolerance ( ) const

inline

Return the tolerance for the quadratic Newton stopping criterion.

Definition at line 263 of file limiter.h.

◆ newton_max_iterations()

template<int dim, typename Number , typename MemorySpace >

DEAL_II_HOST_DEVICE_ALWAYS_INLINE unsigned int ryujin::Euler::LimiterView< dim, Number, MemorySpace >::newton_max_iterations ( ) const

inline

Return the maximal number of quadratic Newton iterations.

Definition at line 272 of file limiter.h.

◆ relaxation_factor()

template<int dim, typename Number , typename MemorySpace >

DEAL_II_HOST_DEVICE_ALWAYS_INLINE ScalarNumber ryujin::Euler::LimiterView< dim, Number, MemorySpace >::relaxation_factor ( ) const

inline

Return the factor used for scaling the relaxation window.

Definition at line 280 of file limiter.h.

◆ projection_bounds_from_state()

template<int dim, typename Number , typename MemorySpace >

DEAL_II_HOST_DEVICE_ALWAYS_INLINE auto ryujin::Euler::LimiterView< dim, Number, MemorySpace >::projection_bounds_from_state	(	const PrecomputedVectorView &	pv,
		const unsigned int	i,
		const state_type &	U_i
	)		const

Given a state U_i and an index i return "strict" bounds, i.e., a minimal convex set containing the state.

Definition at line 418 of file limiter.h.

◆ combine_bounds()

template<int dim, typename Number , typename MemorySpace >

DEAL_II_HOST_DEVICE_ALWAYS_INLINE auto ryujin::Euler::LimiterView< dim, Number, MemorySpace >::combine_bounds	(	const Bounds &	bounds_left,
		const Bounds &	bounds_right
	)		const

Given two bounds bounds_left, bounds_right, this function computes a larger, combined set of bounds that this is a (convex) superset of the two.

Definition at line 433 of file limiter.h.

◆ fully_relax_bounds()

template<int dim, typename Number , typename MemorySpace >

DEAL_II_HOST_DEVICE_ALWAYS_INLINE auto ryujin::Euler::LimiterView< dim, Number, MemorySpace >::fully_relax_bounds	(	const Bounds &	bounds,
		const Number &	hd
	)		const

This function applies a relaxation to a given a (strict) bound bounds using a non dimensionalized measure hd (that should scale as $h^d$, where $h$ is the local mesh size). This is done for the case of the Euler equations by multiplying maximum bounds with $(1+r)$ and minimum bounds with $(1-r)$, while ensuring that the bounds still describe an admissible state.

Definition at line 447 of file limiter.h.

◆ reset()

template<int dim, typename Number , typename MemorySpace >

DEAL_II_HOST_DEVICE_ALWAYS_INLINE void ryujin::Euler::LimiterView< dim, Number, MemorySpace >::reset	(	const PrecomputedVectorView &	pv,
		const unsigned int	i,
		const state_type &	U_i,
		const flux_contribution_type &	flux_i
	)

Reset temporary storage

Definition at line 473 of file limiter.h.

◆ accumulate()

template<int dim, typename Number , typename MemorySpace >

DEAL_II_HOST_DEVICE_ALWAYS_INLINE void ryujin::Euler::LimiterView< dim, Number, MemorySpace >::accumulate	(	const PrecomputedVectorView &	pv,
		const unsigned int *	js,
		const state_type &	U_j,
		const flux_contribution_type &	flux_j,
		const dealii::Tensor< 1, dim, Number > &	scaled_c_ij,
		const state_type &	affine_shift
	)

When looping over the sparsity row, add the contribution associated with the neighboring state U_j.

Definition at line 499 of file limiter.h.

◆ bounds()

template<int dim, typename Number , typename MemorySpace >

DEAL_II_HOST_DEVICE_ALWAYS_INLINE auto ryujin::Euler::LimiterView< dim, Number, MemorySpace >::bounds ( const Number hd_i ) const

Return the computed bounds (with relaxation applied).

Definition at line 550 of file limiter.h.

◆ limit()

template<int dim, typename Number , typename MemorySpace >

DEAL_II_HOST_DEVICE std::tuple< Number, bool > ryujin::Euler::LimiterView< dim, Number, MemorySpace >::limit	(	const Bounds &	bounds,
		const state_type &	U,
		const state_type &	P,
		const Number	t_min = `Number(0.)`,
		const Number	t_max = `Number(1.)`
	)		const

Given a state $\mathbf U$ and an update $\mathbf P$ this function computes and returns the maximal coefficient $t$, obeying $t_{\text{min}} < t < t_{\text{max}}$, such that the selected local minimum principles are obeyed.

The returned boolean is set to true if the original low-order update was within bounds.

Note: If the debug option DEBUG_EXPENSIVE_BOUNDS_CHECK is set to true, then the boolean is set to true if the low-order and the resulting high-order update are within bounds. The latter might be violated due to round-off errors when computing the limiter bounds.

Definition at line 580 of file limiter.h.

References ryujin::compare_and_apply_mask(), ryujin::negative_part(), ryujin::positive_part(), ryujin::pow(), and ryujin::quadratic_newton_step().

Member Data Documentation

◆ problem_dimension

template<int dim, typename Number , typename MemorySpace >

constexpr auto ryujin::Euler::LimiterView< dim, Number, MemorySpace >::problem_dimension = View::problem_dimension

staticconstexpr

Definition at line 217 of file limiter.h.

◆ n_bounds

template<int dim, typename Number , typename MemorySpace >

constexpr unsigned int ryujin::Euler::LimiterView< dim, Number, MemorySpace >::n_bounds = 3

staticconstexpr

The number of stored entries in the bounds array.

Definition at line 235 of file limiter.h.

The documentation for this class was generated from the following file:

source/euler/limiter.h

Public Member Functions

Typedefs and constexpr constants

Computation and manipulation of bounds

Detailed Description

Member Typedef Documentation

◆ View

◆ ScalarNumber

◆ state_type

◆ flux_contribution_type

◆ precomputed_type

◆ PrecomputedVectorView

◆ Bounds

Constructor & Destructor Documentation

◆ LimiterView()

Member Function Documentation

◆ iterations()

◆ newton_tolerance()

◆ newton_max_iterations()

◆ relaxation_factor()

◆ projection_bounds_from_state()

◆ combine_bounds()

◆ fully_relax_bounds()

◆ reset()

◆ accumulate()

◆ bounds()

◆ limit()

Member Data Documentation

◆ problem_dimension

◆ n_bounds