Transport problems¶
Note
These problem types require the transport module and can be used only when that module is included in your OOFEM configuration.
Stationary transport problem¶
Stationary transport problem. A series of loading cases is maintained as a sequence of time steps: for each load case an auxiliary time step is generated whose time equals the load case number, and the load vector for each load case is formed at that auxiliary time.
Record syntax
StationaryProblem nsteps (in)
[sparselinsolverparams (…)]
[exportfields (ia)] [keeptangent]
Parameters
nsteps(in)Number of load cases.
- [
sparselinsolverparams(…)] Attributes of the sparse linear solver; see Sparse linear solver parameters.
- [
exportfields(ia)] See Exporting fields to a subsequent analysis below.
- [
keeptangent] A flag; when present, the tangent is assembled once and reused in every step, which saves time when it does not in fact change.
Exporting fields to a subsequent analysis¶
When the problem is used within a staggered analysis, the field it computes can be exported and made available to subsequent analyses. A temperature field computed here, for instance, can then be taken into account by a subsequent mechanical analysis.
To make that possible the field must be “exported”, which is what the
exportfields array does: it holds the field identifiers under which
the problem registers its primary unknowns. The identifiers are those of the
FieldType enumeration in src/core/field.h. Subsequent analyses can then
access the exported fields, if they support the feature.
Transient transport problem¶
Nonlinear implicit integration scheme for transient transport problems. This is the recommended transient transport solver.
The generalized midpoint rule, sometimes called the \(\alpha\)-method, is used for time discretization, with \(0\le\alpha\le1\):
\(\alpha=0\) gives the explicit Euler forward method;
\(\alpha=0.5\) recovers the implicit trapezoidal rule, which is unconditionally stable and second-order accurate in \(\Delta t\);
\(\alpha=1.0\) gives the implicit Euler backward method, which is unconditionally stable and first-order accurate in \(\Delta t\).
This problem supports sets and changes in the number of equations, so Dirichlet boundary conditions may be imposed and removed during the solution.
Record syntax
TransientTransport nsteps (in)
alpha (rn)
(deltaT (rn) | dTfunction (in) |
prescribedtimes (ra))
[initT (rn)] [lumped]
[keeptangent] [exportfields (ia)]
Parameters
nsteps(in)Number of time steps to solve.
alpha(rn)The \(\alpha\) parameter of the generalized midpoint rule, as above.
deltaT(rn)Time step length used for integration. Give exactly one of
deltaT,dTfunctionorprescribedtimes.dTfunction(in)Number of a time function (see (Time) function records) returning the length of the time step. A variable time step is advantageous when a large time interval has to be covered.
prescribedtimes(ra)Explicit array of solution times.
- [
initT(rn)] Initial time for the integration. Default
0..- [
lumped] A flag; when present, the numerical algorithm is stabilised using a lumped capacity matrix, which reduces the initial oscillations.
- [
keeptangent] A flag; when present, the tangent is assembled once and reused in every step.
- [
exportfields(ia)]
Transient transport problem — linear case (obsolete)¶
Warning
Obsolete. Use Transient transport problem instead; the nonlinear transport solver is strongly preferred for reasons of stability.
Linear implicit integration scheme for transient transport problems, using the same generalized midpoint rule as Transient transport problem.
Record syntax
NonStationaryProblem nsteps (in)
alpha (rn)
(deltaT (rn) | deltaTfunction (in) |
prescribedtimes (ra))
[initT (rn)] [lumpedcapa]
[sparselinsolverparams (…)]
[exportfields (ia)]
[changingProblemSize]
Parameters
nsteps(in)Number of time steps to solve.
alpha(rn)The \(\alpha\) parameter of the generalized midpoint rule.
deltaT(rn)Time step length used for integration.
deltaTfunction(in)Number of a time function (see (Time) function records) returning the length of the time step.
prescribedtimes(ra)Explicit array of solution times.
- [
initT(rn)] Initial time for the integration. Default
0.- [
lumpedcapa] A flag; when present, the numerical algorithm is stabilised using a lumped capacity matrix, which reduces the initial oscillations.
- [
sparselinsolverparams(…)] Attributes of the sparse linear solver; see Sparse linear solver parameters.
- [
exportfields(ia)] - [
changingProblemSize] A flag enabling changes in the number of equations, so that Dirichlet boundary conditions can be imposed and removed during the solution. It works by storing the solution values directly on the nodes (DOFs). If the problem does not grow or shrink during the solution it is more efficient to use the conventional strategy and omit this flag.
Transient transport problem — nonlinear case (obsolete)¶
Warning
Obsolete. Use Transient transport problem instead.
Implicit integration scheme for transient transport problems, using the same generalized midpoint rule as Transient transport problem. See the Material Library Manual for the solution algorithm.
Record syntax
NlTransientTransportProblem nsteps (in)
alpha (rn) rtol (rn)
(deltaT (rn) | deltaTfunction (in))
[initT (rn)] [lumpedcapa]
[nsmax (in)] [manrmsteps (in)]
[sparselinsolverparams (…)]
[exportfields (ia)]
[changingProblemSize]
Parameters
nsteps(in)Number of time steps to solve.
alpha(rn)The \(\alpha\) parameter of the generalized midpoint rule.
rtol(rn)Relative convergence tolerance. The norms of the residual physical quantity — heat or mass — described by the solution vector, and of the change of the solution vector, are evaluated in each iteration; convergence is reached when both are smaller than
rtol.deltaT(rn)Time step length used for integration.
deltaTfunction(in)Number of a time function returning the length of the time step; see Transient transport problem — linear case (obsolete).
- [
initT(rn)] Initial time for the integration; see Transient transport problem — linear case (obsolete).
- [
lumpedcapa] A flag; when present, the numerical algorithm is stabilised using a lumped capacity matrix, which reduces the initial oscillations.
- [
nsmax(in)] Maximum number of iterations per time step. Default
30.- [
manrmsteps(in)] Default
0, for which the modified Newton-Raphson scheme is used and the left-hand side matrix is never updated. A nonzero value switches to the accelerated Newton-Raphson scheme, updating the left-hand side matrix aftermanrmstepssteps.- [
sparselinsolverparams(…)] Attributes of the sparse linear solver; see Sparse linear solver parameters.
- [
exportfields(ia)] - [
changingProblemSize]