***resolution

Description

This procedure is used to fix the calculation parameters and method of resolution for the global matrix problem \(\bf K\bf u=\bf R\). Except for the simplest linear problems, at least some of the procedures sub-options will be required. Non-linear problems are best solved with some substantial customization of these parameters.

Syntax

Specification of the method of resolution requires a definition of the form:

***resolution type

which will be followed by a number of options of the ** command level. The possible types of resolution are summarized in the following table:

Resolution	Description
newton	resolution of type Newton-Raphson, or modified Newton
bfgs	resolution of type BFGS quasi-Newton
riks	a Riks solution to pass by local instabilities or snap-throughs

In the absence of type the default newton will be substituted.

This ***resolution command has a variety of sub-procedures to define load sequencing:

**sequence

Define loading sequences.

**cycles

Define cycles of loading sequences.

**automatic_time

Define automatic time stepping.

**init_d_dof

Accelerate convergence by pre-initializing the DOF increment.

**no_symmetrize

Use the non-symmetric solver.

**max_divergence

Specify the maximum number of divergences allowable before the problem stops.

**skip_cycles

Use the cycle skip algorithm.

**use_lumped_mass

Use the lumped mass matrix (only available for ****calcul dynamics)