Extrapolation

*Extrapolation, <strategy>

Use this keyword to specify an extrapolation strategy of the initial estimate of the solution at the beginning of each increment (except for the first one). By default, a linear extrapolation is performed.

• Theory Manual

  ---

  The theory behind the implementations

You can choose between the following <strategy>:

• linear: a linear extrapolation, i.e. \(\Delta d_{n+1} = \Delta d_n/\Delta t_n \Delta t_{n+1}\)

• quadratic: quadratic extrapolation based on the last three known values of the solution \(\Delta d_{n+1}(d_{n-2},d_{n-1},d_n)\). For the first three increments of each new step, linear extrapolation is used.

• coupled: ...

• constant: an unscaled approximation based on the last increment is assumed, i.e. \(\Delta d_{n+1} = \Delta d_n\).

• none: no extrapolation is performed, the elements are called with \(\Delta d_{n+1} = 0\). This method usually slows down the convergence finding, but can help in case of strong convergence difficulties.