Particles interactions computation. More...

#include "resources/Scripts/types/types.h"
#include "resources/Scripts/KernelFunctions/Kernel.h"

Include dependency graph for MLS.cl:

Macros
#define	__CLEARY__ 10.f

#define	_MLS_ mls[i]

Functions
__kernel void	entry (const __global int imove, const __global int imirrored, const __global vec r, const __global vec rmirrored, const __global float rho, const __global float m, __global matrix mls, __global uint icell, __global uint *ihoc, uint N, uivec4 n_cells)
	Compute the MLS transformation matrix inverse, \( L_i^{-1} \), due to the particles at the other side of the symmetry plane.

Detailed Description

Particles interactions computation.

Macro Definition Documentation

◆ CLEARY

#define __CLEARY__ 10.f

◆ _MLS_

#define _MLS_ mls[i]

Function Documentation

◆ entry()

__kernel void entry	(	const __global int *	imove,
		const __global int *	imirrored,
		const __global vec *	r,
		const __global vec *	rmirrored,
		const __global float *	rho,
		const __global float *	m,
		__global matrix *	mls,
		__global uint *	icell,
		__global uint *	ihoc,
		uint	N,
		uivec4	n_cells
	)

Compute the MLS transformation matrix inverse, \( L_i^{-1} \), due to the particles at the other side of the symmetry plane.

Such transformation matrix can be multiplied by the kernel gradient to produce a new kernel gradient, \( \nabla W^{L}_{ij} = L_i \cdot \nabla W_{ij} \), such that the lienar fields differential operators are consistently computed.

Parameters

imove	Moving flags. imove > 0 for regular fluid particles. imove = 0 for sensors. imove < 0 for boundary elements/particles.
imirrored	0 if the particle has not been mirrored, 1 otherwise.
r	Position \( \mathbf{r} \).
rmirrored	Mirrored position of the particle, r if imirrored is false (0).
rho	Density \( \rho \).
m	Mass \( m \).
mls	Kernel MLS transformation matrix \( L \).
icell	Cell where each particle is located.
ihoc	Head of chain for each cell (first particle found).
N	Number of particles.
n_cells	Number of cells in each direction

Note: The MLS kernel transformation will be computed just for the particles with the moving flag mls_imove, and using just the information of the particles with the moving flag mls_imove

Here is the call graph for this function:

Macros