cs_navsto_param.h File Reference

#include "cs_boundary.h"
#include "cs_cdo_turbulence.h"
#include "cs_equation_param.h"
#include "cs_iter_algo.h"
#include "cs_math.h"
#include "cs_param_sles.h"
#include "cs_physical_constants.h"

Include dependency graph for cs_navsto_param.h:

Go to the source code of this file.

Data Structures
struct	cs_navsto_param_sles_t
	Structure storing the parameters for solving the Navier-Stokes system. More...
struct	cs_navsto_param_boussinesq_t
	Structure storing the parameters related to the Boussinesq source term in the momentum equation. More...
struct	cs_navsto_param_t
	Structure storing the parameters related to the resolution of the Navier-Stokes system. More...

Macros
#define	CS_NAVSTO_STREAM_EQNAME   "streamfunction_eq"

Typedefs
typedef cs_flag_t	cs_navsto_param_model_flag_t
typedef cs_flag_t	cs_navsto_param_post_flag_t

Enumerations
enum	cs_navsto_param_model_t { CS_NAVSTO_MODEL_STOKES, CS_NAVSTO_MODEL_OSEEN, CS_NAVSTO_MODEL_INCOMPRESSIBLE_NAVIER_STOKES, CS_NAVSTO_N_MODELS }
	Describe the system of equations related to the Navier-Stokes to be solved. More...
enum	cs_navsto_param_model_bit_t {   CS_NAVSTO_MODEL_STEADY = 1<<0, CS_NAVSTO_MODEL_GRAVITY_EFFECTS = 1<<1, CS_NAVSTO_MODEL_CORIOLIS_EFFECTS = 1<<2, CS_NAVSTO_MODEL_PASSIVE_THERMAL_TRACER = 1<<3,   CS_NAVSTO_MODEL_BOUSSINESQ = 1<<4, CS_NAVSTO_MODEL_WITH_SOLIDIFICATION = 1<<5 }
	Bit values for additional physical modelling related to the Navier-Stokes system of equations. More...
enum	cs_navsto_param_post_bit_t {   CS_NAVSTO_POST_VELOCITY_DIVERGENCE = 1<< 0, CS_NAVSTO_POST_KINETIC_ENERGY = 1<< 1, CS_NAVSTO_POST_VORTICITY = 1<< 2, CS_NAVSTO_POST_VELOCITY_GRADIENT = 1<< 3,   CS_NAVSTO_POST_STREAM_FUNCTION = 1<< 4, CS_NAVSTO_POST_HELICITY = 1<< 5, CS_NAVSTO_POST_ENSTROPHY = 1<< 6, CS_NAVSTO_POST_MASS_DENSITY = 1<< 7,   CS_NAVSTO_POST_CELL_MASS_FLUX_BALANCE = 1<< 8, CS_NAVSTO_POST_PRESSURE_GRADIENT = 1<< 9 }
	Bit values for additional generic postprocessing related to the Navier-Stokes module. In what follows, w denotes the vorticity vector, u the velocity vector and k the kinetic energy defined by 1/2 * u u. More...
enum	cs_navsto_sles_t {   CS_NAVSTO_SLES_ADDITIVE_GMRES_BY_BLOCK, CS_NAVSTO_SLES_BLOCK_MULTIGRID_CG, CS_NAVSTO_SLES_BY_BLOCKS, CS_NAVSTO_SLES_DIAG_SCHUR_GCR,   CS_NAVSTO_SLES_DIAG_SCHUR_GMRES, CS_NAVSTO_SLES_DIAG_SCHUR_MINRES, CS_NAVSTO_SLES_EQ_WITHOUT_BLOCK, CS_NAVSTO_SLES_GCR,   CS_NAVSTO_SLES_GKB_PETSC, CS_NAVSTO_SLES_GKB_GMRES, CS_NAVSTO_SLES_GKB_SATURNE, CS_NAVSTO_SLES_LOWER_SCHUR_GCR,   CS_NAVSTO_SLES_MINRES, CS_NAVSTO_SLES_MULTIPLICATIVE_GMRES_BY_BLOCK, CS_NAVSTO_SLES_MUMPS, CS_NAVSTO_SLES_NOTAY_TRANSFORM,   CS_NAVSTO_SLES_SGS_SCHUR_GCR, CS_NAVSTO_SLES_UPPER_SCHUR_GCR, CS_NAVSTO_SLES_UPPER_SCHUR_GMRES, CS_NAVSTO_SLES_USER,   CS_NAVSTO_SLES_UZAWA_SCHUR_GCR, CS_NAVSTO_SLES_UZAWA_AL, CS_NAVSTO_SLES_UZAWA_CG, CS_NAVSTO_SLES_N_TYPES }
	High-level information about the way of settings the SLES for solving the Navier-Stokes system. When the system is treated as a saddle-point problem (monolithic approach in what follows), then one uses these notations: A_{00} is the upper-left block and A_{11} (should be 0 but the preconditioner may have entries for the approximation of the inverse of the Schur complement). More...
enum	cs_navsto_param_coupling_t { CS_NAVSTO_COUPLING_ARTIFICIAL_COMPRESSIBILITY, CS_NAVSTO_COUPLING_MONOLITHIC, CS_NAVSTO_COUPLING_PROJECTION, CS_NAVSTO_N_COUPLINGS }
	Choice of algorithm for solving the system. More...
enum	cs_navsto_key_t {   CS_NSKEY_ADVECTION_EXTRAPOL, CS_NSKEY_ADVECTION_FORMULATION, CS_NSKEY_ADVECTION_SCHEME, CS_NSKEY_ADVECTION_STRATEGY,   CS_NSKEY_DOF_REDUCTION, CS_NSKEY_GD_SCALE_COEF, CS_NSKEY_IL_ALGO_ATOL, CS_NSKEY_IL_ALGO_DTOL,   CS_NSKEY_IL_ALGO_RTOL, CS_NSKEY_IL_ALGO_RESTART, CS_NSKEY_IL_ALGO_VERBOSITY, CS_NSKEY_MAX_IL_ALGO_ITER,   CS_NSKEY_MAX_NL_ALGO_ITER, CS_NSKEY_MAX_OUTER_ITER, CS_NSKEY_NL_ALGO, CS_NSKEY_NL_ALGO_ATOL,   CS_NSKEY_NL_ALGO_DTOL, CS_NSKEY_NL_ALGO_RTOL, CS_NSKEY_NL_ALGO_VERBOSITY, CS_NSKEY_QUADRATURE,   CS_NSKEY_SCHUR_STRATEGY, CS_NSKEY_SLES_STRATEGY, CS_NSKEY_SPACE_SCHEME, CS_NSKEY_THERMAL_TOLERANCE,   CS_NSKEY_TIME_SCHEME, CS_NSKEY_TIME_THETA, CS_NSKEY_VERBOSITY, CS_NSKEY_N_KEYS }
	List of available keys for setting the parameters of the Navier-Stokes system. More...

Functions
double	cs_navsto_param_get_notay_scaling (void)
	Retrieve the scaling coefficient used in the Notay's transformation devised in "Algebraic multigrid for Stokes equations" SIAM J. Sci. Comput. Vol. 39 (5), 2017 In this article, this scaling is denoted by alpha. More...
void	cs_navsto_param_set_notay_scaling (double scaling_coef)
	Set the scaling coefficient used in the Notay's transformation devised in "Algebraic multigrid for Stokes equations" SIAM J. Sci. Comput. Vol. 39 (5), 2017 In this article, this scaling is denoted by alpha. More...
cs_navsto_param_t *	cs_navsto_param_create (const cs_boundary_t *boundaries, cs_navsto_param_model_t model, cs_navsto_param_model_flag_t model_flag, cs_navsto_param_coupling_t algo_coupling, cs_navsto_param_post_flag_t post_flag)
	Create a new structure to store all numerical parameters related to the resolution of the Navier-Stokes (NS) system. More...
cs_navsto_param_t *	cs_navsto_param_free (cs_navsto_param_t *param)
	Free a cs_navsto_param_t structure. More...
void	cs_navsto_param_set (cs_navsto_param_t *nsp, cs_navsto_key_t key, const char *keyval)
	Set a parameter attached to a keyname in a cs_navsto_param_t structure. More...
void	cs_navsto_param_transfer (const cs_navsto_param_t *nsp, cs_equation_param_t *eqp)
	Apply the numerical settings defined for the Navier-Stokes system to an equation related to this system. More...
void	cs_navsto_param_log (const cs_navsto_param_t *nsp)
	Summary of the main cs_navsto_param_t structure. More...
cs_navsto_param_boussinesq_t *	cs_navsto_param_add_boussinesq_term (cs_navsto_param_t *nsp, cs_real_t dilatation_coef, cs_real_t reference_value)
	Add a new Boussinesq term (source term for the momemtum equation) More...
void	cs_navsto_param_set_boussinesq_array (cs_navsto_param_boussinesq_t *bp, const cs_real_t *var)
	Set the array of values to consider in the Boussinesq term. More...
cs_navsto_param_sles_t *	cs_navsto_param_get_sles_param (const cs_navsto_param_t *nsp)
	Retrieve the cs_equation_param_t structure related to the velocity equation (momentum equation in most of the cases) More...
cs_equation_param_t *	cs_navsto_param_get_velocity_param (const cs_navsto_param_t *nsp)
	Retrieve the cs_equation_param_t structure related to the velocity equation (momentum equation in most of the cases) More...
const char *	cs_navsto_param_get_model_name (cs_navsto_param_model_t model)
	Retrieve the name of the model system of equations. More...
const char *	cs_navsto_param_get_coupling_name (cs_navsto_param_coupling_t coupling)
	Retrieve the name of the coupling algorithm. More...
void	cs_navsto_set_reference_pressure (cs_navsto_param_t *nsp, cs_real_t pref)
	Set the value to consider for the reference pressure. More...
cs_xdef_t *	cs_navsto_add_velocity_ic_by_value (cs_navsto_param_t *nsp, const char *z_name, cs_real_t *val)
	Define the initial condition for the velocity unknowns. This definition can be done on a specified mesh location. By default, the unknown is set to zero everywhere. Here the initial value is set to a constant value. More...
cs_xdef_t *	cs_navsto_add_velocity_ic_by_analytic (cs_navsto_param_t *nsp, const char *z_name, cs_analytic_func_t *analytic, void *input)
	Define the initial condition for the velocity unknowns. This definition can be done on a specified mesh location. By default, the unknown is set to zero everywhere. Here the initial value is set according to an analytical function. More...
cs_xdef_t *	cs_navsto_add_pressure_ic_by_value (cs_navsto_param_t *nsp, const char *z_name, cs_real_t *val)
	Define the initial condition for the pressure unknowns. This definition can be done on a specified mesh location. By default, the unknown is set to zero everywhere. Here the initial value is set to a constant value. More...
cs_xdef_t *	cs_navsto_add_pressure_ic_by_analytic (cs_navsto_param_t *nsp, const char *z_name, cs_analytic_func_t *analytic, void *input)
	Define the initial condition for the pressure unknowns. This definition can be done on a specified mesh location. By default, the unknown is set to zero everywhere. Here the initial value is set according to an analytical function. More...
void	cs_navsto_set_fixed_walls (cs_navsto_param_t *nsp)
	Add the definition of boundary conditions related to a fixed wall into the set of parameters for the management of the Navier-Stokes system of equations. More...
void	cs_navsto_set_symmetries (cs_navsto_param_t *nsp)
	Add the definition of boundary conditions related to a symmetry into the set of parameters for the management of the Navier-Stokes system of equations. More...
void	cs_navsto_set_outlets (cs_navsto_param_t *nsp)
	Add the definition of boundary conditions related to outlets into the set of parameters for the management of the Navier-Stokes system of equations. More...
cs_xdef_t *	cs_navsto_set_pressure_bc_by_value (cs_navsto_param_t *nsp, const char *z_name, cs_real_t *values)
	Set the pressure field on a boundary using a uniform value. More...
cs_xdef_t *	cs_navsto_set_velocity_wall_by_value (cs_navsto_param_t *nsp, const char *z_name, cs_real_t *values)
	Define the velocity field for a sliding wall boundary using a uniform value. More...
cs_xdef_t *	cs_navsto_set_velocity_inlet_by_value (cs_navsto_param_t *nsp, const char *z_name, cs_real_t *values)
	Define the velocity field for an inlet boundary using a uniform value. More...
cs_xdef_t *	cs_navsto_set_velocity_inlet_by_analytic (cs_navsto_param_t *nsp, const char *z_name, cs_analytic_func_t *ana, void *input)
	Define the velocity field for an inlet boundary using an analytical function. More...
cs_xdef_t *	cs_navsto_set_velocity_inlet_by_array (cs_navsto_param_t *nsp, const char *z_name, cs_flag_t loc, cs_real_t *array, bool is_owner, cs_lnum_t *index)
	Define the velocity field for an inlet boundary using an array of values. More...
cs_xdef_t *	cs_navsto_set_velocity_inlet_by_dof_func (cs_navsto_param_t *nsp, const char *z_name, cs_dof_func_t *func, void *func_input)
	Define the velocity field for an inlet boundary using a DoF function. More...
cs_xdef_t *	cs_navsto_add_source_term_by_analytic (cs_navsto_param_t *nsp, const char *z_name, cs_analytic_func_t *ana, void *input)
	Define a new source term structure defined by an analytical function. More...
cs_xdef_t *	cs_navsto_add_source_term_by_val (cs_navsto_param_t *nsp, const char *z_name, cs_real_t *val)
	Define a new source term structure defined by a constant value. More...
cs_xdef_t *	cs_navsto_add_source_term_by_array (cs_navsto_param_t *nsp, const char *z_name, cs_flag_t loc, cs_real_t *array, bool is_owner, cs_lnum_t *index)
	Define a new source term structure defined by an array. More...
void	cs_navsto_add_oseen_field (cs_navsto_param_t *nsp, cs_adv_field_t *adv_fld)
	Add a advection field for the Oseen problem. More...

Macro Definition Documentation

CS_NAVSTO_STREAM_EQNAME

#define CS_NAVSTO_STREAM_EQNAME   "streamfunction_eq"

Typedef Documentation

cs_navsto_param_model_flag_t

typedef cs_flag_t cs_navsto_param_model_flag_t

cs_navsto_param_post_flag_t

typedef cs_flag_t cs_navsto_param_post_flag_t

Enumeration Type Documentation

cs_navsto_key_t

enum cs_navsto_key_t

List of available keys for setting the parameters of the Navier-Stokes system.

Enumerator
CS_NSKEY_ADVECTION_EXTRAPOL	Set the extrapolation to use for the estimation of the advection field (cf. cs_param_advection_extrapol_t))
CS_NSKEY_ADVECTION_FORMULATION	Set the type of formulation for the advection term, for example in the Oseen problem. (cf. cs_param_advection_form_t)
CS_NSKEY_ADVECTION_SCHEME	Set the type of scheme for the advection term, for example in the Oseen problem. (cf. cs_param_advection_scheme_t)
CS_NSKEY_ADVECTION_STRATEGY	Set the strategy to handle the advection term (cf. cs_param_advection_strategy_t)
CS_NSKEY_DOF_REDUCTION	Set how the DoFs are defined (similar to CS_EQKEY_DOF_REDUCTION) Enable to set this type of DoFs definition for all related equations
CS_NSKEY_GD_SCALE_COEF	Set the scaling of the grad-div term when an artificial compressibility algorithm or an Uzawa-Augmented Lagrangian method is used
CS_NSKEY_IL_ALGO_ATOL	Absolute tolerance at which the Oseen or Stokes system is resolved. These systems corresponds to an inner linear system to solve when considering the Navier-Stokes system since one has to handle the non-linearity in addition as an outer process.
CS_NSKEY_IL_ALGO_DTOL	Divergence tolerance at which the Oseen or Stokes system is resolved. These systems corresponds to an inner linear system to solve when considering the Navier-Stokes system since one has to handle the non-linearity in addition as an outer process.
CS_NSKEY_IL_ALGO_RTOL	Relative tolerance at which the Oseen or Stokes system is resolved. These systems corresponds to an inner linear system to solve when considering the Navier-Stokes system since one has to handle the non-linearity in addition as an outer process.
CS_NSKEY_IL_ALGO_RESTART	Number of iterations before restarting a Krylov solver as the main solver (useful if the strategy implied a GMRES, flexible GMRES or GCR)
CS_NSKEY_IL_ALGO_VERBOSITY	Level of verbosity related to the inner linear algorithm (cf. CS_NSKEY_SLES_STRATEGY)
CS_NSKEY_MAX_IL_ALGO_ITER	Set the maximal number of iteration for solving the inner linear system.
CS_NSKEY_MAX_NL_ALGO_ITER	Set the maximal number of iterations for solving the non-linearity arising from the advection form
CS_NSKEY_MAX_OUTER_ITER	Set the maximal number of outer iterations for solving the full system including the turbulence modelling or the thermal system for instance
CS_NSKEY_NL_ALGO	Type of algorithm to consider to solve the non-linearity arising from the Navier-Stokes system (Picard or Anderson)
CS_NSKEY_NL_ALGO_ATOL	Absolute tolerance at which the non-linearity arising from the advection term is resolved
CS_NSKEY_NL_ALGO_DTOL	Divergence tolerance at which the non-linearity arising from the advection term is detected
CS_NSKEY_NL_ALGO_RTOL	Relative tolerance at which the non-linearity arising from the advection term is resolved
CS_NSKEY_NL_ALGO_VERBOSITY	Level of verbosity related to the non-linear algorithm
CS_NSKEY_QUADRATURE	Set the type to use in all functions involving quadrature (similar to CS_EQKEY_BC_QUADRATURE)
CS_NSKEY_SCHUR_STRATEGY	Set the way to define the Schur complement approximation (cf. cs_param_schur_approx_t)
CS_NSKEY_SLES_STRATEGY	Strategy for solving the SLES arising from the discretization of the Navier-Stokes system
CS_NSKEY_SPACE_SCHEME	Numerical scheme for the space discretization. Available choices are: "cdo_fb" or "cdofb" for CDO face-based scheme
CS_NSKEY_THERMAL_TOLERANCE	Value of the tolerance criterion under which one stops iterating on the Navier-Stokes and thermal systems
CS_NSKEY_TIME_SCHEME	Numerical scheme for the time discretization
CS_NSKEY_TIME_THETA	Set the value of theta. Only useful if CS_NSKEY_TIME_SCHEME is set to "theta_scheme" Example: "0.75" (keyval must be between 0 and 1)
CS_NSKEY_VERBOSITY	Set the level of details for the specific part related to the Navier-Stokes system
CS_NSKEY_N_KEYS

cs_navsto_param_coupling_t

enum cs_navsto_param_coupling_t

Choice of algorithm for solving the system.

Enumerator
CS_NAVSTO_COUPLING_ARTIFICIAL_COMPRESSIBILITY	The system is solved using an artificial compressibility algorithm. One vectorial equation is solved followed by a pressure update.
CS_NAVSTO_COUPLING_MONOLITHIC	The system is treated as a "monolithic" matrix
CS_NAVSTO_COUPLING_PROJECTION	The system is solved using an incremental projection algorithm
CS_NAVSTO_N_COUPLINGS

cs_navsto_param_model_bit_t

enum cs_navsto_param_model_bit_t

Bit values for additional physical modelling related to the Navier-Stokes system of equations.

Enumerator
CS_NAVSTO_MODEL_STEADY	There is no time dependency in the model
CS_NAVSTO_MODEL_GRAVITY_EFFECTS	Take into account the gravity effects (add a constant source term equal to rho*vect(g))
CS_NAVSTO_MODEL_CORIOLIS_EFFECTS	Take into account the Coriolis effects (add a source term)
CS_NAVSTO_MODEL_PASSIVE_THERMAL_TRACER	An additional equation is created involving the thermal equation. The advection field is automatically set as the mass flux. By default, the temperature is solved in Celsius and homogeneous Neumann boundary conditions (no flux) are set. This option is not compatible with the option CS_NAVSTO_MODEL_BOUSSINESQ
CS_NAVSTO_MODEL_BOUSSINESQ	An additional equation is created involving the thermal equation. The advection field is automatically set as the mass flux. By default, the temperature is solved in Celsius and homogeneous Neumann boundary conditions (no flux) are set. The variation of mass density around a reference value is related to the variation of temperature w.r.t. a reference temperature and it plays a role in the gravity effects (rho.vect(g)) The gradient of temperature is assumed to have a small norm and the mass density variates in a small range. In this case, an additional momentum source term is added.
CS_NAVSTO_MODEL_WITH_SOLIDIFICATION	The Navier-Stokes is modified to take into account solidification process. A boussinesq term is added as well as a head loss term derived from a Darcy approximation. Please see the cs_solidification_model_t structure and related structures for more details

cs_navsto_param_model_t

enum cs_navsto_param_model_t

Describe the system of equations related to the Navier-Stokes to be solved.

Enumerator
CS_NAVSTO_MODEL_STOKES	Stokes equations (mass and momentum) with the classical choice of variables i.e. velocity and pressure. Mass density is assumed to be constant.
CS_NAVSTO_MODEL_OSEEN	Like the incompressible Navier-Stokes equations (mass and momentum) but with a velocity field which is given. Thus the advection term in the momentum equation is linear. Unknowns: velocity and pressure. Mass density is assumed to be constant. The advection field is set using cs_navsto_add_oseen_field
CS_NAVSTO_MODEL_INCOMPRESSIBLE_NAVIER_STOKES	Navier-Stokes equations: mass and momentum with a constant mass density Mass equation is equivalent to an incompressibility constraint
CS_NAVSTO_N_MODELS

cs_navsto_param_post_bit_t

enum cs_navsto_param_post_bit_t

Bit values for additional generic postprocessing related to the Navier-Stokes module. In what follows, w denotes the vorticity vector, u the velocity vector and k the kinetic energy defined by 1/2 * u u.

Enumerator
CS_NAVSTO_POST_VELOCITY_DIVERGENCE	Compute div(u) and associate a field to this quantity
CS_NAVSTO_POST_KINETIC_ENERGY	Compute rho/2 u u and associate a field to this quantity
CS_NAVSTO_POST_VORTICITY	Compute w = curl(u) and associate a field to this quantity
CS_NAVSTO_POST_VELOCITY_GRADIENT	Compute the tensor grad(u) and associate a field to this quantity
CS_NAVSTO_POST_STREAM_FUNCTION	Add an equation to compute the stream function and associate a field to this quantity. This equation is a scalar valued diffusion equation: -Lap(Psi) = w_z This is only useful for a 2D computation (assuming that the z-azis is the extruded one, i.e. the flow is in the x-y plane
CS_NAVSTO_POST_HELICITY	Compute h = u w and associate a field to this quantity
CS_NAVSTO_POST_ENSTROPHY	Compute w w and associate a field to this quantity
CS_NAVSTO_POST_MASS_DENSITY	Compute the mass density in each cell and monitor the evolution of the integral of the mass in the full computational domain. If a Boussinesq approximation is set, then the mass density used in the buoyancy term is considered (not the reference value assumed to be constant).
CS_NAVSTO_POST_CELL_MASS_FLUX_BALANCE	Compute the balance of the mass flux for each cell. When the mass density is constant, this quantity is only a scaling (by the mass density) of the quantity computed with the flag CS_NAVSTO_POST_VELOCITY_DIVERGENCE.
CS_NAVSTO_POST_PRESSURE_GRADIENT	Compute the presure gradient in each cell.

cs_navsto_sles_t

enum cs_navsto_sles_t

High-level information about the way of settings the SLES for solving the Navier-Stokes system. When the system is treated as a saddle-point problem (monolithic approach in what follows), then one uses these notations: A_{00} is the upper-left block and A_{11} (should be 0 but the preconditioner may have entries for the approximation of the inverse of the Schur complement).

Enumerator
CS_NAVSTO_SLES_ADDITIVE_GMRES_BY_BLOCK	Associated keyword: "additive_gmres" Available choice when a monolithic approach is used (i.e. with the parameter CS_NAVSTO_COUPLING_MONOLITHIC is set as coupling algorithm) The Navier-Stokes system of equations is solved an additive preconditioner (block diagonal matrix where the block 00 is A_{00}) and the block 11 is set to the identity. Preconditioner/solver for the block 00 is set using the momentum equation. This option is only available with the support to the PETSc library up to now.
CS_NAVSTO_SLES_BLOCK_MULTIGRID_CG	Associated keyword: "block_amg_cg" The Navier-Stokes system of equations is solved using a multigrid on each diagonal block as a preconditioner and applying a conjugate gradient as solver. Use this strategy when the saddle-point problem has been reformulated into a "classical" linear system. For instance when a Uzawa or an Artificial Compressibility coupling algorithm is used. (i.e. with the parameter CS_NAVSTO_COUPLING_ARTIFICIAL_COMPRESSIBILITY is set as coupling algorithm). This option is only available with the support to the PETSc library up to now.
CS_NAVSTO_SLES_BY_BLOCKS
CS_NAVSTO_SLES_DIAG_SCHUR_GCR	Associated keyword: "diag_schur_gcr" The Stokes or Navier-Stokes system is solved using a GCR algorithm with a block diagonal preconditioner using a Schur approximation for the pressure block (the block 22). The system is stored using a hybrid assembled/unassembled blocks. The velocity block is assembled (with potentially sub-blocks for each component) and the velocity divergence/pressure gradient operators are unassembled.
CS_NAVSTO_SLES_DIAG_SCHUR_GMRES	Associated keyword: "diag_schur_gmres" Available choice when a monolithic approach is used (i.e. with the parameter CS_NAVSTO_COUPLING_MONOLITHIC is set as coupling algorithm). The Navier-Stokes system of equations is solved using a block diagonal preconditioner where the block 00 is A_{00} preconditioned with one multigrid iteration and the block 11 is an approximation of the Schur complement preconditioned with one multigrid iteration. The main iterative solver is a flexible GMRES. This option is only available with the support to the PETSc library up to now.
CS_NAVSTO_SLES_DIAG_SCHUR_MINRES	Associated keyword: "diag_schur_minres" The Stokes or Navier-Stokes system with an explicit advection is solved using a MINRES algorithm with a block diagonal preconditioner using a Schur approximation for the pressure block (the block 22). The system is stored using a hybrid assembled/unassembled blocks. The velocity block is assembled (with potentially sub-blocks for each component) and the velocity divergence/pressure gradient operators are unassembled.
CS_NAVSTO_SLES_EQ_WITHOUT_BLOCK	Associated keyword: "no_block" Use the same mechanism as for a stand-alone equation. In this case, the setting relies on the function cs_equation_set_sles and the different options for solving a linear system such as the choice of the iterative solver or the choice of the preconditioner or the type of residual normalization
CS_NAVSTO_SLES_GCR	Associated keyword: "gcr" The Stokes or Navier-Stokes system is solved using a GCR algorithm without preconditioning. The system is stored using a hybrid assembled/unassembled blocks. The velocity block is assembled (with potentially sub-blocks for each component) and the velocity divergence/pressure gradient operators are unassembled.
CS_NAVSTO_SLES_GKB_PETSC	Associated keyword: "gkb_petsc" Available choice when a monolithic approach is used (i.e. with the parameter CS_NAVSTO_COUPLING_MONOLITHIC is set as coupling algorithm). The Navier-Stokes system of equations is solved using a Golub-Kahan bi-diagonalization. One assumes that the saddle-point system is symmetric. By default, the block A_{00} may be augmented (this is not the default choice) and is solved with a conjugate gradient algorithm preconditioned with a multigrid. The residual is computed in the energy norm. This option is only available with the support to the PETSc library up to now.
CS_NAVSTO_SLES_GKB_GMRES	Associated keyword: "gkb_gmres" Available choice when a monolithic approach is used (i.e. with the parameter CS_NAVSTO_COUPLING_MONOLITHIC is set as coupling algorithm). The Navier-Stokes system of equations is solved using a Golub-Kahan bi-diagonalization (GKB) as preconditioner of a flexible GMRES solver. The GKB algorithm is solved with a reduced tolerance as well as the CG+Multigrid used as an inner solver in the GKB algorithm. One assumes that the saddle-point system is symmetric. The residual for the GKB part is computed in the energy norm. This option is only available with the support to the PETSc library up to now.
CS_NAVSTO_SLES_GKB_SATURNE	Associated keyword: "gkb" or "gkb_saturne" Available choice when a monolithic approach is used (i.e. with the parameter CS_NAVSTO_COUPLING_MONOLITHIC is set as coupling algorithm). The Navier-Stokes system of equations is solved using a Golub-Kahan bi-diagonalization. By default, the block A_{00} may be augmented (this is not the default choice) and is solved with the SLES settings given to the momentum equation A conjugate gradient algorithm preconditioned with a multigrid for Stokes for instance. The residual is computed in the energy norm.
CS_NAVSTO_SLES_LOWER_SCHUR_GCR	Associated keyword: "lower_schur_gcr" The Stokes or Navier-Stokes system is solved using a GCR algorithm with an lower triangular block preconditioner using a Schur approximation for the pressure block (the block 22). The system is stored using a hybrid assembled/unassembled blocks. The velocity block is assembled (with potentially sub-blocks for each component) and the velocity divergence/pressure gradient operators are unassembled.
CS_NAVSTO_SLES_MINRES	Associated keyword: "minres" The Stokes or Navier-Stokes system with an explicit advection is solved using a MINRES algorithm without preconditioning. The system is stored using a hybrid assembled/unassembled blocks. The velocity block is assembled (with potentially sub-blocks for each component) and the velocity divergence/pressure gradient operators are unassembled.
CS_NAVSTO_SLES_MULTIPLICATIVE_GMRES_BY_BLOCK	Associated keyword: "multiplicative_gmres" Available choice when a monolithic approach is used (i.e. with the parameter CS_NAVSTO_COUPLING_MONOLITHIC is set as coupling algorithm) The Navier-Stokes system of equations is solved a multiplicative preconditioner (block diagonal matrix where the block 00 is A_{00}) and the block 11 is set to the identity. Block 01 is also considered in the block preconditioner. Preconditioner/solver for the block 00 is set using the momentum equation. This option is only available with the support to the PETSc library up to now.
CS_NAVSTO_SLES_MUMPS	Associated keyword: "mumps" Direct solver to solve the full (saddle-point) system arising from the discretization of the Navier-Stokes equations
CS_NAVSTO_SLES_NOTAY_TRANSFORM	Associated keyword: "notay" Transform the saddle-point problem into an equivalent system without a zero block for the (2,2) block. This transformation allows one to consider standard preconditionner and iterative Krylow solver.
CS_NAVSTO_SLES_SGS_SCHUR_GCR	Associated keyword: "sgs_schur_gcr" The Stokes or Navier-Stokes system is solved using a GCR algorithm with a symmetric Gauss-Seidel block preconditioner using a Schur approximation for the pressure block (the block 22). The system is stored using a hybrid assembled/unassembled blocks. The velocity block is assembled (with potentially sub-blocks for each component) and the velocity divergence/pressure gradient operators are unassembled.
CS_NAVSTO_SLES_UPPER_SCHUR_GCR	Associated keyword: "upper_schur_gcr" The Stokes or Navier-Stokes system is solved using a GCR algorithm with an upper triangular block preconditioner using a Schur approximation for the pressure block (the block 22). The system is stored using a hybrid assembled/unassembled blocks. The velocity block is assembled (with potentially sub-blocks for each component) and the velocity divergence/pressure gradient operators are unassembled.
CS_NAVSTO_SLES_UPPER_SCHUR_GMRES	Associated keyword: "upper_schur_gmres" Available choice when a monolithic approach is used (i.e. with the parameter CS_NAVSTO_COUPLING_MONOLITHIC is set as coupling algorithm). The Navier-Stokes system of equations is solved using a upper triangular block preconditioner where the block 00 is A_{00} preconditioned with one multigrid iteration and the block 11 is an approximation of the Schur complement preconditioned with a minres. The main iterative solver is a flexible GMRES. This option is only available with the support to the PETSc library up to now.
CS_NAVSTO_SLES_USER	Associated keyword: "user" Resolution using a user-defined function. This function fulfills a pre-defined prototype
CS_NAVSTO_SLES_UZAWA_SCHUR_GCR	Associated keyword: "uza_schur_gcr" The Stokes or Navier-Stokes system is solved using a GCR algorithm with an Uzawa algorithm tuned for block preconditioning and using a Schur approximation for the pressure block (the block 22). The system is stored using a hybrid assembled/unassembled blocks. The velocity block is assembled (with potentially sub-blocks for each component) and the velocity divergence/pressure gradient operators are unassembled.
CS_NAVSTO_SLES_UZAWA_AL	Associated keyword: "uzawa_al" Resolution using an uzawa algorithm with an Augmented Lagrangian approach
CS_NAVSTO_SLES_UZAWA_CG	Associated keyword: "uzawa_cg" Resolution using an uzawa algorithm optimized using a conjugate gradient reformulation. Two systems are solved at each iteration (one related to the velocity block, one related to the Schur complement approximation - size of the pressure space).
CS_NAVSTO_SLES_N_TYPES

Function Documentation

cs_navsto_add_oseen_field()

void cs_navsto_add_oseen_field	(	cs_navsto_param_t *	nsp,
		cs_adv_field_t *	adv_fld
	)

Add a advection field for the Oseen problem.

Parameters

[in,out]	nsp	pointer to a cs_navsto_param_t
[in,out]	adv_fld	pointer to a cs_adv_field_t

cs_navsto_add_pressure_ic_by_analytic()

cs_xdef_t* cs_navsto_add_pressure_ic_by_analytic	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_analytic_func_t *	analytic,
		void *	input
	)

Define the initial condition for the pressure unknowns. This definition can be done on a specified mesh location. By default, the unknown is set to zero everywhere. Here the initial value is set according to an analytical function.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" if all cells are considered)
[in]	analytic	pointer to an analytic function
[in]	input	NULL or pointer to a structure cast on-the-fly

Returns

a pointer to the new cs_xdef_t structure

cs_navsto_add_pressure_ic_by_value()

cs_xdef_t* cs_navsto_add_pressure_ic_by_value	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_real_t *	val
	)

Define the initial condition for the pressure unknowns. This definition can be done on a specified mesh location. By default, the unknown is set to zero everywhere. Here the initial value is set to a constant value.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" if all cells are considered)
[in]	val	pointer to the value

Returns

a pointer to the new cs_xdef_t structure

cs_navsto_add_source_term_by_analytic()

cs_xdef_t* cs_navsto_add_source_term_by_analytic	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_analytic_func_t *	ana,
		void *	input
	)

Define a new source term structure defined by an analytical function.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" all cells are considered)
[in]	ana	pointer to an analytical function
[in]	input	NULL or pointer to a structure cast on-the-fly

Returns

a pointer to the new cs_xdef_t structure

cs_navsto_add_source_term_by_array()

cs_xdef_t* cs_navsto_add_source_term_by_array	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_flag_t	loc,
		cs_real_t *	array,
		bool	is_owner,
		cs_lnum_t *	index
	)

Define a new source term structure defined by an array.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" all cells are considered)
[in]	loc	information to know where are located values
[in]	array	pointer to an array
[in]	is_owner	transfer the lifecycle to the cs_xdef_t structure (true or false)
[in]	index	optional pointer to the array index

Returns

a pointer to the new cs_xdef_t structure

cs_navsto_add_source_term_by_val()

cs_xdef_t* cs_navsto_add_source_term_by_val	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_real_t *	val
	)

Define a new source term structure defined by a constant value.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" all cells are considered)
[in]	val	pointer to the value to set

Returns

a pointer to the new cs_xdef_t structure

cs_navsto_add_velocity_ic_by_analytic()

cs_xdef_t* cs_navsto_add_velocity_ic_by_analytic	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_analytic_func_t *	analytic,
		void *	input
	)

Define the initial condition for the velocity unknowns. This definition can be done on a specified mesh location. By default, the unknown is set to zero everywhere. Here the initial value is set according to an analytical function.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" if all cells are considered)
[in]	analytic	pointer to an analytic function
[in]	input	NULL or pointer to a structure cast on-the-fly

Returns

a pointer to the new cs_xdef_t structure

cs_navsto_add_velocity_ic_by_value()

cs_xdef_t* cs_navsto_add_velocity_ic_by_value	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_real_t *	val
	)

Define the initial condition for the velocity unknowns. This definition can be done on a specified mesh location. By default, the unknown is set to zero everywhere. Here the initial value is set to a constant value.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" if all cells are considered)
[in]	val	pointer to the value

Returns

a pointer to the new cs_xdef_t structure

cs_navsto_param_add_boussinesq_term()

cs_navsto_param_boussinesq_t* cs_navsto_param_add_boussinesq_term	(	cs_navsto_param_t *	nsp,
		cs_real_t	dilatation_coef,
		cs_real_t	reference_value
	)

Add a new Boussinesq term (source term for the momemtum equation)

Parameters

[in,out]

nsp

pointer to a cs_navsto_param_t structure

Returns

a pointer to the newly added structure

cs_navsto_param_create()

cs_navsto_param_t* cs_navsto_param_create	(	const cs_boundary_t *	boundaries,
		cs_navsto_param_model_t	model,
		cs_navsto_param_model_flag_t	model_flag,
		cs_navsto_param_coupling_t	algo_coupling,
		cs_navsto_param_post_flag_t	post_flag
	)

Create a new structure to store all numerical parameters related to the resolution of the Navier-Stokes (NS) system.

Parameters

[in]	boundaries	pointer to a cs_boundary_t structure
[in]	model	type of model related to the NS system
[in]	model_flag	additional high-level model options
[in]	algo_coupling	algorithm used for solving the NS system
[in]	post_flag	predefined post-processings options

Returns

a pointer to a new allocated structure

cs_navsto_param_free()

cs_navsto_param_t* cs_navsto_param_free

(

cs_navsto_param_t *

param

)

Free a cs_navsto_param_t structure.

Parameters

[in,out]

param

pointer to a cs_navsto_param_t structure

Returns

a NULL pointer

cs_navsto_param_get_coupling_name()

const char* cs_navsto_param_get_coupling_name

(

cs_navsto_param_coupling_t

coupling

)

Retrieve the name of the coupling algorithm.

Parameters

[in]

coupling

a cs_navsto_param_coupling_t

Returns

the name

cs_navsto_param_get_model_name()

const char* cs_navsto_param_get_model_name

(

cs_navsto_param_model_t

model

)

Retrieve the name of the model system of equations.

Parameters

[in]

model

a cs_navsto_param_model_t

Returns

the corresponding name

cs_navsto_param_get_notay_scaling()

double cs_navsto_param_get_notay_scaling

(

void

)

Retrieve the scaling coefficient used in the Notay's transformation devised in "Algebraic multigrid for Stokes equations" SIAM J. Sci. Comput. Vol. 39 (5), 2017 In this article, this scaling is denoted by alpha.

Returns

the value of the scaling coefficient

cs_navsto_param_get_sles_param()

cs_navsto_param_sles_t* cs_navsto_param_get_sles_param

(

const cs_navsto_param_t *

nsp

)

Retrieve the cs_equation_param_t structure related to the velocity equation (momentum equation in most of the cases)

Parameters

[in]

nsp

pointer to a cs_navsto_param_t structure

Returns

a pointer to the set of SLES parameters

cs_navsto_param_get_velocity_param()

cs_equation_param_t* cs_navsto_param_get_velocity_param

(

const cs_navsto_param_t *

nsp

)

Retrieve the cs_equation_param_t structure related to the velocity equation (momentum equation in most of the cases)

Parameters

[in]

nsp

pointer to a cs_navsto_param_t structure

Returns

a pointer to the set of parameters related to the momentum equation

cs_navsto_param_log()

void cs_navsto_param_log

(

const cs_navsto_param_t *

nsp

)

Summary of the main cs_navsto_param_t structure.

Parameters

[in]

nsp

pointer to a cs_navsto_param_t structure

cs_navsto_param_set()

void cs_navsto_param_set	(	cs_navsto_param_t *	nsp,
		cs_navsto_key_t	key,
		const char *	keyval
	)

Set a parameter attached to a keyname in a cs_navsto_param_t structure.

Parameters

[in,out]	nsp	pointer to a cs_navsto_param_t structure to set
[in]	key	key related to the member of eq to set
[in]	keyval	accessor to the value to set

cs_navsto_param_set_boussinesq_array()

void cs_navsto_param_set_boussinesq_array	(	cs_navsto_param_boussinesq_t *	bp,
		const cs_real_t *	var
	)

Set the array of values to consider in the Boussinesq term.

Parameters

[in,out]	bp	pointer to a cs_navsto_param_boussinesq_t structure
[in]	var	shared pointer to the array of values to consider

cs_navsto_param_set_notay_scaling()

void cs_navsto_param_set_notay_scaling

(

double

scaling_coef

)

Set the scaling coefficient used in the Notay's transformation devised in "Algebraic multigrid for Stokes equations" SIAM J. Sci. Comput. Vol. 39 (5), 2017 In this article, this scaling is denoted by alpha.

Parameters

[in]

scaling_coef

valued of the scaling coefficient

cs_navsto_param_transfer()

void cs_navsto_param_transfer	(	const cs_navsto_param_t *	nsp,
		cs_equation_param_t *	eqp
	)

Apply the numerical settings defined for the Navier-Stokes system to an equation related to this system.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in,out]	eqp	pointer to a cs_equation_param_t structure

cs_navsto_set_fixed_walls()

void cs_navsto_set_fixed_walls

(

cs_navsto_param_t *

nsp

)

Add the definition of boundary conditions related to a fixed wall into the set of parameters for the management of the Navier-Stokes system of equations.

Parameters

[in]

nsp

pointer to a cs_navsto_param_t structure

cs_navsto_set_outlets()

void cs_navsto_set_outlets

(

cs_navsto_param_t *

nsp

)

Add the definition of boundary conditions related to outlets into the set of parameters for the management of the Navier-Stokes system of equations.

Parameters

[in]

nsp

pointer to a cs_navsto_param_t structure

cs_navsto_set_pressure_bc_by_value()

cs_xdef_t* cs_navsto_set_pressure_bc_by_value	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_real_t *	values
	)

Set the pressure field on a boundary using a uniform value.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" all boundary faces are considered)
[in]	value	value to set

Returns

a pointer to the new cs_xdef_t structure

cs_navsto_set_reference_pressure()

void cs_navsto_set_reference_pressure	(	cs_navsto_param_t *	nsp,
		cs_real_t	pref
	)

Set the value to consider for the reference pressure.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	pref	value of the reference pressure

cs_navsto_set_symmetries()

void cs_navsto_set_symmetries

(

cs_navsto_param_t *

nsp

)

Add the definition of boundary conditions related to a symmetry into the set of parameters for the management of the Navier-Stokes system of equations.

Parameters

[in]

nsp

pointer to a cs_navsto_param_t structure

cs_navsto_set_velocity_inlet_by_analytic()

cs_xdef_t* cs_navsto_set_velocity_inlet_by_analytic	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_analytic_func_t *	ana,
		void *	input
	)

Define the velocity field for an inlet boundary using an analytical function.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" all boundary faces are considered)
[in]	ana	pointer to an analytical function
[in]	input	NULL or pointer to a structure cast on-the-fly

Returns

a pointer to the new cs_xdef_t structure

cs_navsto_set_velocity_inlet_by_array()

cs_xdef_t* cs_navsto_set_velocity_inlet_by_array	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_flag_t	loc,
		cs_real_t *	array,
		bool	is_owner,
		cs_lnum_t *	index
	)

Define the velocity field for an inlet boundary using an array of values.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" all boundary faces are considered)
[in]	loc	information to know where are located values
[in]	array	pointer to an array
[in]	is_owner	transfer the lifecycle to the cs_xdef_t structure (true or false)
[in]	index	optional pointer to the array index

Returns

a pointer to the new cs_xdef_t structure

cs_navsto_set_velocity_inlet_by_dof_func()

cs_xdef_t* cs_navsto_set_velocity_inlet_by_dof_func	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_dof_func_t *	func,
		void *	func_input
	)

Define the velocity field for an inlet boundary using a DoF function.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" all boundary faces are considered)
[in]	func	pointer to a cs_dof_function_t
[in]	func_input	NULL or pointer to a structure cast on-the-fly

Returns

a pointer to the new cs_xdef_t structure

cs_navsto_set_velocity_inlet_by_value()

cs_xdef_t* cs_navsto_set_velocity_inlet_by_value	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_real_t *	values
	)

Define the velocity field for an inlet boundary using a uniform value.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" all boundary faces are considered)
[in]	values	array of three real values

Returns

a pointer to the new cs_xdef_t structure

cs_navsto_set_velocity_wall_by_value()

cs_xdef_t* cs_navsto_set_velocity_wall_by_value	(	cs_navsto_param_t *	nsp,
		const char *	z_name,
		cs_real_t *	values
	)

Define the velocity field for a sliding wall boundary using a uniform value.

Parameters

[in]	nsp	pointer to a cs_navsto_param_t structure
[in]	z_name	name of the associated zone (if NULL or "" all boundary faces are considered)
[in]	values	array of three real values

Returns

a pointer to the new cs_xdef_t structure