Solution Operation Nodes and Solvers
The first few sections provide some background information about the solvers and the algorithms used:
•
Selecting a Stationary, Time-Dependent, or Eigenvalue Solver
•
Remarks on Solver-Related Model Characteristics
•
Scaling of Variables and Equations
•
About the Stationary Solver
•
About the Parametric Solver
•
About the Time-Dependent Solver
•
About the Time-Discrete Solver
•
Termination Criterion for the Fully Coupled and Segregated Attribute Nodes
•
The Eigenvalue Solver Algorithms
•
The Modal Solver Algorithm
•
The Time-Explicit Algorithms
About Solver Commands in the COMSOL Multiphysics Programming Reference Manual
The settings for the solver operation nodes listed in Table 20-6 are detailed. There is also a list of the References for the Solution Operation Nodes and Solvers.
Table 20-6: Solution Operation Nodes.
Icon
Name
Description
AWE Solver
Solve a parametric problem with asymptotic waveform evaluation.
Dependent Variables
Handles the dependent variables solved for (initial values, scaling) and dependent variables not solved for (prescribed values).
Eigenvalue Solver
Solve linear or linearized eigenvalue problems (also called eigenfrequency problems). Also see The Eigenvalue Solver Algorithms.
FFT Solver
Use a forward FFT to transform a time-dependent solution to the frequency domain or an inverse NFT/FFT to transform a frequency-domain solution to the time domain.
Modal Reduction
To solve either parameter stepping (also called frequency response) or time stepping (also called transient response) problems using a reduced model. This approach also produces a reduced-order model or reduced matrices. Also see The Modal Solver Algorithm.
Modal Solver
To solve either parameter stepping (also called frequency response) or time stepping (also called transient response) problems using a reduced model and a modal approach. Also see The Modal Solver Algorithm.
Optimization Solver
Solve PDE-constrained optimization problems. Requires the Optimization Module.
POD Reduction
Model reduction using the POD (proper orthogonal decomposition) method.
Plug Flow Solver
Solve a plug flow reactor model (requires the Chemical Reaction Engineering Module).
Stationary Solver
Solve linear and nonlinear stationary problems (also called static or steady-state problems). Also see About the Stationary Solver.
Time-Dependent Solver
Solve time-dependent problems (also called dynamic or unsteady problems) using the implicit BDF or generalized-α methods or one of the available explicit Runge–Kutta, Adams–Bashforth 3, and Verlet types of time-stepping methods. Also see About the Time-Dependent Solver.
Time-Discrete Solver
Solve time-dependent problems (dynamic or unsteady problems) that have already been discretized in time using, for example, the prev operator or the bdf operator.