MFEM
v4.1.0
Finite element discretization library
|
Interface to the CVODE library – linear multi-step methods. More...
#include <sundials.hpp>
Public Member Functions | |
CVODESolver (int lmm) | |
Construct a serial wrapper to SUNDIALS' CVODE integrator. More... | |
CVODESolver (MPI_Comm comm, int lmm) | |
Construct a parallel wrapper to SUNDIALS' CVODE integrator. More... | |
void | Init (TimeDependentOperator &f_) |
Initialize CVODE: calls CVodeCreate() to create the CVODE memory and set some defaults. More... | |
virtual void | Step (Vector &x, double &t, double &dt) |
Integrate the ODE with CVODE using the specified step mode. More... | |
void | UseMFEMLinearSolver () |
Attach the linear system setup and solve methods from the TimeDependentOperator i.e., SUNImplicitSetup() and SUNImplicitSolve() to CVODE. More... | |
void | UseSundialsLinearSolver () |
Attach SUNDIALS GMRES linear solver to CVODE. More... | |
void | SetStepMode (int itask) |
Select the CVODE step mode: CV_NORMAL (default) or CV_ONE_STEP. More... | |
void | SetSStolerances (double reltol, double abstol) |
Set the scalar relative and scalar absolute tolerances. More... | |
void | SetMaxStep (double dt_max) |
Set the maximum time step. More... | |
void | SetMaxOrder (int max_order) |
Set the maximum method order. More... | |
void | PrintInfo () const |
Print various CVODE statistics. More... | |
virtual | ~CVODESolver () |
Destroy the associated CVODE memory and SUNDIALS objects. More... | |
Public Member Functions inherited from mfem::ODESolver | |
ODESolver () | |
virtual void | Run (Vector &x, double &t, double &dt, double tf) |
Perform time integration from time t [in] to time tf [in]. More... | |
virtual | ~ODESolver () |
Public Member Functions inherited from mfem::SundialsSolver | |
void * | GetMem () const |
Access the SUNDIALS memory structure. More... | |
int | GetFlag () const |
Returns the last flag retured by a call to a SUNDIALS function. More... | |
Static Protected Member Functions | |
static int | RHS (realtype t, const N_Vector y, N_Vector ydot, void *user_data) |
Wrapper to compute the ODE rhs function. More... | |
static int | LinSysSetup (realtype t, N_Vector y, N_Vector fy, SUNMatrix A, booleantype jok, booleantype *jcur, realtype gamma, void *user_data, N_Vector tmp1, N_Vector tmp2, N_Vector tmp3) |
Setup the linear system \( A x = b \). More... | |
static int | LinSysSolve (SUNLinearSolver LS, SUNMatrix A, N_Vector x, N_Vector b, realtype tol) |
Solve the linear system \( A x = b \). More... | |
Protected Attributes | |
int | lmm_type |
Linear multistep method type. More... | |
int | step_mode |
CVODE step mode (CV_NORMAL or CV_ONE_STEP). More... | |
Protected Attributes inherited from mfem::ODESolver | |
TimeDependentOperator * | f |
Pointer to the associated TimeDependentOperator. More... | |
MemoryType | mem_type |
Protected Attributes inherited from mfem::SundialsSolver | |
void * | sundials_mem |
SUNDIALS mem structure. More... | |
int | flag |
Last flag returned from a call to SUNDIALS. More... | |
bool | reinit |
Flag to signal memory reinitialization is need. More... | |
long | saved_global_size |
Global vector length on last initialization. More... | |
N_Vector | y |
State vector. More... | |
SUNMatrix | A |
Linear system A = I - gamma J, M - gamma J, or J. More... | |
SUNMatrix | M |
Mass matrix M. More... | |
SUNLinearSolver | LSA |
Linear solver for A. More... | |
SUNLinearSolver | LSM |
Linear solver for M. More... | |
SUNNonlinearSolver | NLS |
Nonlinear solver. More... | |
Additional Inherited Members | |
Protected Member Functions inherited from mfem::SundialsSolver | |
bool | Parallel () const |
bool | Parallel () const |
SundialsSolver () | |
Protected constructor: objects of this type should be constructed only as part of a derived class. More... | |
Static Protected Attributes inherited from mfem::SundialsSolver | |
static constexpr double | default_rel_tol = 1e-4 |
Default scalar relative tolerance. More... | |
static constexpr double | default_abs_tol = 1e-9 |
Default scalar absolute tolerance. More... | |
Interface to the CVODE library – linear multi-step methods.
Definition at line 92 of file sundials.hpp.
mfem::CVODESolver::CVODESolver | ( | int | lmm | ) |
Construct a serial wrapper to SUNDIALS' CVODE integrator.
[in] | lmm | Specifies the linear multistep method, the options are:
|
Definition at line 120 of file sundials.cpp.
mfem::CVODESolver::CVODESolver | ( | MPI_Comm | comm, |
int | lmm | ||
) |
Construct a parallel wrapper to SUNDIALS' CVODE integrator.
[in] | comm | The MPI communicator used to partition the ODE system |
[in] | lmm | Specifies the linear multistep method, the options are:
|
Definition at line 129 of file sundials.cpp.
|
virtual |
Destroy the associated CVODE memory and SUNDIALS objects.
Definition at line 405 of file sundials.cpp.
|
virtual |
Initialize CVODE: calls CVodeCreate() to create the CVODE memory and set some defaults.
If the CVODE memory has already been created, it checks if the problem size has changed since the last call to Init(). If the problem is the same then CVodeReInit() will be called in the next call to Step(). If the problem size has changed, the CVODE memory is freed and realloced for the new problem size.
[in] | f_ | The TimeDependentOperator that defines the ODE system. |
Reimplemented from mfem::ODESolver.
Definition at line 151 of file sundials.cpp.
|
staticprotected |
Setup the linear system \( A x = b \).
Definition at line 94 of file sundials.cpp.
|
staticprotected |
Solve the linear system \( A x = b \).
Definition at line 109 of file sundials.cpp.
void mfem::CVODESolver::PrintInfo | ( | ) | const |
Print various CVODE statistics.
Definition at line 360 of file sundials.cpp.
|
staticprotected |
Wrapper to compute the ODE rhs function.
Definition at line 78 of file sundials.cpp.
void mfem::CVODESolver::SetMaxOrder | ( | int | max_order | ) |
Set the maximum method order.
CVODE uses adaptive-order integration, based on the local truncation error. The default values for max_order are 12 for CV_ADAMS and 5 for CV_BDF. Use this if you know a priori that your system is such that higher order integration formulas are unstable.
Definition at line 354 of file sundials.cpp.
void mfem::CVODESolver::SetMaxStep | ( | double | dt_max | ) |
Set the maximum time step.
Definition at line 348 of file sundials.cpp.
void mfem::CVODESolver::SetSStolerances | ( | double | reltol, |
double | abstol | ||
) |
Set the scalar relative and scalar absolute tolerances.
Definition at line 342 of file sundials.cpp.
void mfem::CVODESolver::SetStepMode | ( | int | itask | ) |
Select the CVODE step mode: CV_NORMAL (default) or CV_ONE_STEP.
[in] | itask | The desired step mode. |
Definition at line 337 of file sundials.cpp.
|
virtual |
Integrate the ODE with CVODE using the specified step mode.
[in,out] | x | On output, the solution vector at the requested output time tout = t + dt. |
[in,out] | t | On output, the output time reached. |
[in,out] | dt | On output, the last time step taken. |
Implements mfem::ODESolver.
Definition at line 256 of file sundials.cpp.
void mfem::CVODESolver::UseMFEMLinearSolver | ( | ) |
Attach the linear system setup and solve methods from the TimeDependentOperator i.e., SUNImplicitSetup() and SUNImplicitSolve() to CVODE.
Definition at line 291 of file sundials.cpp.
void mfem::CVODESolver::UseSundialsLinearSolver | ( | ) |
Attach SUNDIALS GMRES linear solver to CVODE.
Definition at line 322 of file sundials.cpp.
|
protected |
Linear multistep method type.
Definition at line 95 of file sundials.hpp.
|
protected |
CVODE step mode (CV_NORMAL or CV_ONE_STEP).
Definition at line 96 of file sundials.hpp.