mod_compute_cell_advection_term_exp_semi_lagrangian Module Reference

Explicit semi lagrangian method for the advection equation. More...

Functions/Subroutines
subroutine	compute_cell_advection_term_exp_semi_lagrangian (velocity, time_step, scalar, flux_type, directions, divut)
	Fills divuT with \(u \cdot \nabla T\) with a Semi-Lagrangian method.

Detailed Description

Be aware that this is not a conservative scheme but a traditional advection scheme.

The equation solved is:

\[ \frac{D \phi}{D t} = \frac{\partial \phi}{\partial t} + u \cdot \nabla \phi = 0 \]

in a lagrangian way, using advection of particles and interpolation.

Function/Subroutine Documentation

compute_cell_advection_term_exp_semi_lagrangian()

subroutine mod_compute_cell_advection_term_exp_semi_lagrangian::compute_cell_advection_term_exp_semi_lagrangian	(	type(t_face_field), intent(in)	velocity,
		double precision, intent(in)	time_step,
		double precision, dimension(:,:,:), intent(in)	scalar,
		type(t_fv_flux), intent(in)	flux_type,
		integer, dimension(:), intent(in)	directions,
		double precision, dimension(:,:,:), intent(inout)	divut )

Parameters

[in]	velocity	the velocity u
[in]	time_step	the time step \( \delta t = t^{n+1} - t^n \)
[in]	scalar	the scalar T to be advected
[in]	flux_type	(unused, but necessary for genericity)
[in]	directions	the vector made of directions ([1,2,3] for all of them or [1] for only "x"). This is used for splitted methods.
[in,out]	divut	the resulting field

Todo

MCO: 3rd dimension

MCO: deal with boundaries

MCO: correct with reconstruction of the field

MCO: question the precision with Eulerian vs RK schemes

MCO: deal with splitted method (with "directions")