#include <flsubdom.h>
Public Member Functions | |
flsubdom () | |
void | initiation (gmatrix *gm, FILE *out) |
void | solve_lin_alg_system (double *lhs, double *rhs, FILE *out) |
~flsubdom () | |
Public Attributes | |
double * | compli |
compliances | |
double * | lambda |
long | nch |
the number of changes in the vector of Lagrange multipliers | |
double * | totlambda |
class flsubdom serves for problems with floating subdomains solved by the FETI method
JK, 20.6.2006, modified 3. 8. 2012
Definition at line 16 of file flsubdom.h.
flsubdom | ( | ) |
void initiation | ( | gmatrix * | gm, | |
FILE * | out | |||
) |
function initializes variables and arrays
JK, 3. 8. 2012
Definition at line 30 of file flsubdom.cpp.
References slesolv::feti, seqfeti::matrices_assembl(), Mp, and probdesc::ssle.
void solve_lin_alg_system | ( | double * | lhs, | |
double * | rhs, | |||
FILE * | out | |||
) |
function solves system of linear algebraic equations by the modified conjugate gradient method Lagrange multipliers are computed first and then displacements are obtained for more details see Kruis: Domain Decomposition Methods for Distributed Computing. Saxe-Coburg, 2006.
lhs | - left hand side vector (vector of nodal unknowns) | |
rhs | - right hand side |
JK, 20.6.2006, modified 2.8.2012
Definition at line 48 of file flsubdom.cpp.
References seqfeti::assemble_ff(), seqfeti::evector(), slesolv::feti, Mp, seqfeti::mpcg(), seqfeti::nodalunknowns(), and probdesc::ssle.
double* compli |
double* lambda |
the vector of Lagrange multipliers it contains increments of multipliers in nonlinear problems
Definition at line 27 of file flsubdom.h.
Referenced by flsubdom(), and ~flsubdom().
long nch |
the number of changes in the vector of Lagrange multipliers
Definition at line 33 of file flsubdom.h.
Referenced by flsubdom().
double* totlambda |
the vector of total Lagrange multipliers it is used only in nonlinear problems
Definition at line 30 of file flsubdom.h.
Referenced by flsubdom(), and ~flsubdom().