FEHeatSolver Class Reference

#include <FEHeatSolver.h>

Inheritance diagram for FEHeatSolver:

Public Member Functions
	FEHeatSolver (FEM &fem)
	constructor
bool	Init ()
	one-time initialization
bool	SolveStep (double time)
	solve a timestep
void	Serialize (DumpFile &ar)
	serialize data
bool	InitEquations ()
	Initialize linear equation system.
void	AssembleStiffness (matrix &ke, vector< int > &lm)
	assemble element stiffness matrix
Data Fields
vector< double >	m_T
	temperature array
vector< double >	m_Tp
	previous temperatures
vector< double >	m_R
	residual
vector< double >	m_u
	prescribed temperatures
Protected Member Functions
void	Residual ()
	calculate the residual
bool	StiffnessMatrix ()
	calculate the stiffness matrix
bool	ReformStiffness ()
	form the stiffness matrix
void	Update (vector< double > &u)
	update solution

---

Detailed Description

The FEHeatSolver solves heat-conduction problems

---

Constructor & Destructor Documentation

FEHeatSolver::FEHeatSolver

(

FEM &

fem

)

constructor

constructor for the class

---

Member Function Documentation

void FEHeatSolver::AssembleStiffness	(	matrix &	ke,
		vector< int > &	lm
	)

assemble element stiffness matrix

Assembles the element stiffness matrix into the global stiffness matrix. This function also modifies the residual according to the prescribed degrees of freedom.

bool FEHeatSolver::Init

(

)

[virtual]

one-time initialization

Do one-time initialization for data.

Reimplemented from FESolver.

bool FEHeatSolver::InitEquations

(

)

[virtual]

Initialize linear equation system.

This function initializes the equation system. It is assumed that all free dofs up until now have been given an ID >= 0 and the fixed or rigid dofs an ID < 0. After this operation the nodal ID array will contain the equation number assigned to the corresponding degree of freedom. To distinguish between free or unconstrained dofs and constrained ones the following rules apply to the ID array:

/ | >= 0 --> dof j of node i is a free dof ID[i][j] < == -1 --> dof j of node i is a fixed (no equation assigned too) | < -1 --> dof j of node i is constrained and has equation nr = -ID[i][j]-2 \

Implements FESolver.

bool FEHeatSolver::ReformStiffness

(

)

[protected]

form the stiffness matrix

Reform the stiffness matrix. That is, calculate the shape of the stiffness matrix (i.e. figure out the sparsity pattern), fill the stiffness matrix with the element contributions and then factor the matrix.

void FEHeatSolver::Residual

(

)

[protected]

calculate the residual

Calculate the residual.

void FEHeatSolver::Serialize

(

DumpFile &

ar

)

[virtual]

serialize data

Serializes data to the archive. Still need to implement this.

Implements FESolver.

bool FEHeatSolver::SolveStep

(

double

time

)

[virtual]

solve a timestep

solve a time step

Implements FESolver.

bool FEHeatSolver::StiffnessMatrix

(

)

[protected]

calculate the stiffness matrix

Calculate the global stiffness matrix. This function simply calls FEHeatSolver::ElementStiffness() for each element and then assembles the element stiffness matrix into the global matrix.

---

The documentation for this class was generated from the following files:

• FEBio/FEHeatSolver.h
• FEBio/FEHeatSolver.cpp