petsc-3.7.7 2017-09-25

PetscDSSetJacobianPreconditioner

Set the pointwise Jacobian preconditioner function for given test and basis fields. If this is missing, the system matrix is used to build the preconditioner.

Synopsis

#include "petscds.h" 
PetscErrorCode PetscDSSetJacobianPreconditioner(PetscDS prob, PetscInt f, PetscInt g,
                                  void (*g0)(PetscInt dim, PetscInt Nf, PetscInt NfAux,
                                             const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[],
                                             const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[],
                                             PetscReal t, PetscReal u_tShift, const PetscReal x[], PetscScalar g0[]),
                                  void (*g1)(PetscInt dim, PetscInt Nf, PetscInt NfAux,
                                             const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[],
                                             const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[],
                                             PetscReal t, PetscReal u_tShift, const PetscReal x[], PetscScalar g1[]),
                                  void (*g2)(PetscInt dim, PetscInt Nf, PetscInt NfAux,
                                             const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[],
                                             const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[],
                                             PetscReal t, PetscReal u_tShift, const PetscReal x[], PetscScalar g2[]),
                                  void (*g3)(PetscInt dim, PetscInt Nf, PetscInt NfAux,
                                             const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[],
                                             const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[],
                                             PetscReal t, PetscReal u_tShift, const PetscReal x[], PetscScalar g3[]))

Not collective

Input Parameters

	prob	- The PetscDS
	f	- The test field number
	g	- The field number
	g0	- integrand for the test and basis function term
	g1	- integrand for the test function and basis function gradient term
	g2	- integrand for the test function gradient and basis function term
	g3	- integrand for the test function gradient and basis function gradient term

Note: We are using a first order FEM model for the weak form

\int_\Omega \phi g_0(u, u_t, \nabla u, x, t) \psi + \phi {\vec g}_1(u, u_t, \nabla u, x, t) \nabla \psi + \nabla\phi \cdot {\vec g}_2(u, u_t, \nabla u, x, t) \psi + \nabla\phi \cdot {\overleftrightarrow g}_3(u, u_t, \nabla u, x, t) \cdot \nabla \psi

The calling sequence for the callbacks g0, g1, g2 and g3 is given by

g0(PetscInt dim, PetscInt Nf, PetscInt NfAux,

   const PetscInt uOff[], const PetscInt uOff_x[], const PetscScalar u[], const PetscScalar u_t[], const PetscScalar u_x[],

   const PetscInt aOff[], const PetscInt aOff_x[], const PetscScalar a[], const PetscScalar a_t[], const PetscScalar a_x[],

   PetscReal t, const PetscReal x[], PetscScalar g0[])

	dim	- the spatial dimension
	Nf	- the number of fields
	uOff	- the offset into u[] and u_t[] for each field
	uOff_x	- the offset into u_x[] for each field
	u	- each field evaluated at the current point
	u_t	- the time derivative of each field evaluated at the current point
	u_x	- the gradient of each field evaluated at the current point
	aOff	- the offset into a[] and a_t[] for each auxiliary field
	aOff_x	- the offset into a_x[] for each auxiliary field
	a	- each auxiliary field evaluated at the current point
	a_t	- the time derivative of each auxiliary field evaluated at the current point
	a_x	- the gradient of auxiliary each field evaluated at the current point
	t	- current time
	u_tShift	- the multiplier a for dF/dU_t
	x	- coordinates of the current point
	g0	- output values at the current point

PetscDSSetJacobianPreconditioner

Synopsis

Input Parameters

Note: We are using a first order FEM model for the weak form

The calling sequence for the callbacks g0, g1, g2 and g3 is given by

See Also