scalar_transport.hpp File Reference

flow — cell-centred scalar transport (advection–diffusion) on the cut-cell grid. More...

#include <Kokkos_Core.hpp>
#include <string>
#include "mac_cutcell.hpp"
#include "staggered_advection.hpp"

Go to the source code of this file.

Classes
struct	peclet::flow::ScalarField

Namespaces
namespace	peclet
namespace	peclet::flow

Enumerations
enum class	peclet::flow::ScalarBc { peclet::flow::Periodic = 0 , peclet::flow::Neumann = 1 , peclet::flow::Dirichlet = 2 }

Functions
void	peclet::flow::scalarBuildDiffusionOpen (CCField AC, CCField AW, CCField AE, CCField AS, CCField AN, CCField AB, CCField AT, CCConst ox, CCConst oy, CCConst oz, double D, double idt, C3 e, int g)
void	peclet::flow::scalarBuildRhs (CCField b, CCConst cOld, CCConst U, CCConst V, CCConst W, CCConst ox, CCConst oy, CCConst oz, double idt, int scheme, C3 e, int g)

Detailed Description

flow — cell-centred scalar transport (advection–diffusion) on the cut-cell grid.

A transported scalar c (temperature, concentration, phase fraction) obeys, in the solver's divided-by-dt convention (dx = 1 grid units; physical diffusivity converted by the Python layer):

(1/dt)(c^{n+1} - c^n) + div(open u c) = div(open D grad c) + S

Diffusion is backward-Euler implicit (an openness-weighted 7-band operator, solved by the same red-black Gauss–Seidel as the pressure Poisson — cutcellSmoothColor); advection is explicit, conservative flux-form, reusing the momentum limiter helpers (sadv::tvd / sou / fou_flux) with the MAC face-normal velocities (staggered: C[fd].u is the -fd face velocity, co-located with the face openness). Closed faces (openness 0) carry no flux and no diffusion, so an immersed solid is adiabatic (zero-flux) for free and solid cells stay frozen (A_C = 1/dt, all off-diagonals 0).

This header holds the field-agnostic kernels + the per-scalar state; the Solver owns the scalars and calls advanceScalars() at the end of step() with the just-projected divergence-free velocity.

Definition in file scalar_transport.hpp.