Convergence Criteria

A reliable convergence criterion is essential for any iterative solution procedure. It serves as a fundamental benchmark to evaluate whether the iterative process is approaching a satisfactory solution. The convergence criteria include:

  • Displacement norm

    • The required accuracy is measured by the Euclidean displacement norm (default).

    • The default value is 1e-6

  • Energy norm

    • The required accuracy is measured by the energy norm. This value if not used if the convergence norm is set to Displacement.

    • The default value is 1e-6

  • Absolute convergence

    • The required accuracy is measured by the the incremental norm between two iterations. The absolute convergence limit is applied to terminate the iteration when the change is found sufficiently small (i.e. such as load group with zero current).

    • The iterations are interrupted if the accuracy is smaller than a value which a function of the diameter and the line length scaled by a small number.

For details see the theory manual and input reference, Iteration procedure and Input reference.

The absolute convergence is not a user input.