GSPH force computation using Riemann solver results. More...

#include "shambase/constants.hpp"
#include "shambackends/math.hpp"
#include "shambackends/sycl.hpp"
#include "shammodels/gsph/math/reconstruction.hpp"
#include "shammodels/gsph/math/riemann/iterative.hpp"
#include "shammodels/sph/math/forces.hpp"

Include dependency graph for forces.hpp:

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

Go to the source code of this file.

Namespaces
namespace	shammodels
	namespace for models

Functions
template<class Tvec, class Tscal>
Tscal	shammodels::gsph::gsph_energy_rate (Tscal m_b, Tscal p_star, Tscal v_star, Tscal rho_a_sq, Tscal rho_b_sq, Tscal omega_a, Tscal omega_b, Tvec v_a, Tvec r_ab_unit, Tvec nabla_W_a, Tvec nabla_W_b)
	Compute GSPH energy equation contribution.
template<class Tvec, class Tscal>
void	shammodels::gsph::add_gsph_force_contribution (Tscal m_b, Tscal p_star, Tscal v_star, Tscal rho_a, Tscal rho_b, Tscal omega_a, Tscal omega_b, Tscal Fab_a, Tscal Fab_b, Tvec r_ab_unit, Tvec v_a, Tvec &dv_dt, Tscal &du_dt)
	Add GSPH force contribution from a single neighbor pair.
template<class Tvec, class Tscal>
void	shammodels::gsph::add_gsph_force_contribution_inutsuka (Tscal m_b, Tscal p_star, Tscal v_star, Tscal V2_ij, Tvec grad_W_ij, Tvec r_ab_unit, Tvec v_a, Tvec &dv_dt, Tscal &du_dt)
	Add Inutsuka (2002) GSPH force contribution from a single neighbor pair.
template<class Kernel, class Tvec, class Tscal>
void	shammodels::gsph::accumulate_gsph_pair_force (bool use_inutsuka_v2, Tscal pmass, Tscal p_star, Tscal v_star, Tscal rho_a, Tscal rho_b, Tscal omega_a, Tscal omega_b, Tscal rab, Tscal rab_inv, Tscal h_a, Tscal h_b, Tvec r_ab_unit, Tvec vxyz_a, Tvec &sum_axyz, Tscal &sum_du_a)
	Dispatch a single neighbor pair's force contribution to ChaWhitworth or InutsukaV2, given the pair's Riemann solver result.

Detailed Description

GSPH force computation using Riemann solver results.

Author: Guo Yansong (guo.y.nosp@m.anso.nosp@m.ng.ng.nosp@m.y@gm.nosp@m.ail.c.nosp@m.om); Timothée David–Cléris (tim.s.nosp@m.hamr.nosp@m.ock@p.nosp@m.roto.nosp@m.n.me); Yona Lapeyre (yona..nosp@m.lape.nosp@m.yre@e.nosp@m.ns-l.nosp@m.yon.f.nosp@m.r)

Implements the Godunov SPH (GSPH) force formulation following Cha & Whitworth (2003). The key difference from standard SPH is that the interface pressure p* comes from solving the Riemann problem, rather than using artificial viscosity.

References:

Cha, S.-H. & Whitworth, A.P. (2003) "Implementations and tests of Godunov-type particle hydrodynamics"
Inutsuka, S. (2002) "Reformulation of Smoothed Particle Hydrodynamics with Riemann Solver"

Definition in file forces.hpp.

Function Documentation

◆ accumulate_gsph_pair_force()

template<class Kernel, class Tvec, class Tscal>

void shammodels::gsph::accumulate_gsph_pair_force	(	bool	use_inutsuka_v2,
		Tscal	pmass,
		Tscal	p_star,
		Tscal	v_star,
		Tscal	rho_a,
		Tscal	rho_b,
		Tscal	omega_a,
		Tscal	omega_b,
		Tscal	rab,
		Tscal	rab_inv,
		Tscal	h_a,
		Tscal	h_b,
		Tvec	r_ab_unit,
		Tvec	vxyz_a,
		Tvec &	sum_axyz,
		Tscal &	sum_du_a )

inline

Dispatch a single neighbor pair's force contribution to ChaWhitworth or InutsukaV2, given the pair's Riemann solver result.

Shared by update_derivs_iterative()/update_derivs_exact() in UpdateDerivs.cpp, which only differ in which Riemann solver produced (p_star, v_star).

Template Parameters

Kernel	SPH kernel type (e.g. SPHKernel<Tscal>), for Kernel::dW_3d
Tvec	Vector type
Tscal	Scalar type

Parameters

	use_inutsuka_v2	Selects InutsukaV2 (true) or ChaWhitworth (false)
	pmass	Particle mass (equal-mass GSPH)
	p_star	Interface pressure from Riemann solver
	v_star	Interface velocity from Riemann solver
	rho_a	Density of particle a
	rho_b	Density of particle b
	omega_a	Grad-h correction factor for particle a
	omega_b	Grad-h correction factor for particle b
	rab	Pair separation \|r_a - r_b\|
	rab_inv	Inverse of rab
	h_a	Smoothing length of particle a
	h_b	Smoothing length of particle b
	r_ab_unit	Unit vector from a to b
	vxyz_a	Velocity of particle a
[out]	sum_axyz	Accumulated acceleration
[out]	sum_du_a	Accumulated energy rate

Definition at line 232 of file forces.hpp.

◆ add_gsph_force_contribution()

template<class Tvec, class Tscal>

void shammodels::gsph::add_gsph_force_contribution	(	Tscal	m_b,
		Tscal	p_star,
		Tscal	v_star,
		Tscal	rho_a,
		Tscal	rho_b,
		Tscal	omega_a,
		Tscal	omega_b,
		Tscal	Fab_a,
		Tscal	Fab_b,
		Tvec	r_ab_unit,
		Tvec	v_a,
		Tvec &	dv_dt,
		Tscal &	du_dt )

inline

Add GSPH force contribution from a single neighbor pair.

Convenience function that computes both acceleration and energy rate contributions from a single particle pair, given the Riemann solver result.

Template Parameters

Tvec	Vector type
Tscal	Scalar type

Parameters

	m_b	Mass of neighbor particle
	p_star	Interface pressure from Riemann solver
	v_star	Interface velocity from Riemann solver
	rho_a	Density of particle a
	rho_b	Density of particle b
	omega_a	Grad-h correction factor for particle a
	omega_b	Grad-h correction factor for particle b
	Fab_a	Kernel gradient magnitude \|nabla W_ab(h_a)\|
	Fab_b	Kernel gradient magnitude \|nabla W_ab(h_b)\|
	r_ab_unit	Unit vector from a to b (points toward b)
	v_a	Velocity of particle a
[out]	dv_dt	Accumulated acceleration
[out]	du_dt	Accumulated energy rate

Definition at line 121 of file forces.hpp.

◆ add_gsph_force_contribution_inutsuka()

template<class Tvec, class Tscal>

void shammodels::gsph::add_gsph_force_contribution_inutsuka	(	Tscal	m_b,
		Tscal	p_star,
		Tscal	v_star,
		Tscal	V2_ij,
		Tvec	grad_W_ij,
		Tvec	r_ab_unit,
		Tvec	v_a,
		Tvec &	dv_dt,
		Tscal &	du_dt )

inline

Add Inutsuka (2002) GSPH force contribution from a single neighbor pair.

Uses the effective volume/face formulation instead of the Cha & Whitworth symmetric SPH form: acc -= m_b * p* * V2_ij * grad_W_ij, where V2_ij is the effective squared volume element between the pair (see math/reconstruction.hpp) and grad_W_ij is the pair-symmetrized kernel gradient.

Template Parameters

Tvec	Vector type
Tscal	Scalar type

Parameters

	m_b	Mass of neighbor particle
	p_star	Interface pressure from Riemann solver
	v_star	Interface velocity from Riemann solver
	V2_ij	Effective squared volume element for the pair
	grad_W_ij	Pair-symmetrized kernel gradient vector
	r_ab_unit	Unit vector from a to b (points toward b)
	v_a	Velocity of particle a
[out]	dv_dt	Accumulated acceleration
[out]	du_dt	Accumulated energy rate

Definition at line 187 of file forces.hpp.

◆ gsph_energy_rate()

template<class Tvec, class Tscal>

Tscal shammodels::gsph::gsph_energy_rate	(	Tscal	m_b,
		Tscal	p_star,
		Tscal	v_star,
		Tscal	rho_a_sq,
		Tscal	rho_b_sq,
		Tscal	omega_a,
		Tscal	omega_b,
		Tvec	v_a,
		Tvec	r_ab_unit,
		Tvec	nabla_W_a,
		Tvec	nabla_W_b )

inline

Compute GSPH energy equation contribution.

Following Cha & Whitworth (2003), the energy equation uses the same symmetric force as the momentum equation: f_ab = m_b * p* * (nabla_W_a / (rho_a^2 * Omega_a) + nabla_W_b / (rho_b^2 * Omega_b)) du_a/dt = -f_ab dot (v* - v_a)

This ensures proper energy conservation in shocks by using the same force that appears in the momentum equation.

Template Parameters

Tvec	Vector type
Tscal	Scalar type

Parameters

m_b	Mass of particle b
p_star	Interface pressure from Riemann solver
v_star	Interface velocity (scalar, in direction of r_ab)
rho_a_sq	Density squared of particle a
rho_b_sq	Density squared of particle b
omega_a	Grad-h correction factor for particle a
omega_b	Grad-h correction factor for particle b
v_a	Velocity of particle a
r_ab_unit	Unit vector from a to b
nabla_W_a	Kernel gradient at r_ab with smoothing length h_a
nabla_W_b	Kernel gradient at r_ab with smoothing length h_b

Returns: Energy rate contribution from this pair

Definition at line 70 of file forces.hpp.

Here is the call graph for this function:

Namespaces

Functions

Detailed Description

Function Documentation

◆ accumulate_gsph_pair_force()

◆ add_gsph_force_contribution()

◆ add_gsph_force_contribution_inutsuka()

◆ gsph_energy_rate()