Application | PLAXIS 2D |

Version | PLAXIS 2D |

Date created | 16 April 2012 |

Date modified | 16 April 2012 |

The time step used by Plaxis in a dynamic calculation is constant and equal to* δt = Δt/ (m* n)*, where *Δt* is the Time interval specified for the relevant phase, *m* is the number of Additional steps and *n* is the number of Dynamic sub steps. The result of the multiplication of the Additional step number (m) and the Dynamic sub steps number (n) gives the total number of steps to be used in the time discretization. It is important to define a proper number of steps such that the dynamic signal used in dynamic loading is properly covered. In general it is recommended to choose *Δt*, *m* and *n* in such a way that the dynamic sub step time interval *δt* is equal to the time interval used in the input signal.

The number of the additional steps specifies the number of the steps which can be used in plots in the Output program. A higher number of Additional steps provides more detailed plots, however, the processing time required by the Output program is increased as well. For each given number of additional time step, PLAXIS estimates the number of sub steps on the basis of the generated mesh and the calculated *δt _{critical}* (see theory on critical time, Section 5.2.1 of the Scientific Manual). If the wave velocities (functions of material stiffness) in a model exhibit remarkable differences and/or the model contains very small elements, the standard number of sub steps can be very large. In such situations it may not always be vital to follow the automatic time stepping with the standard number of dynamic sub steps. It is possible to change the calculated number of Dynamic sub steps in the Manual settings window in the Iterative procedure box of the Parameters tabsheet. Changing the number of dynamic sub steps will also influence the time step (

When the time step per dynamic substep (*δt*) differs from the time step size in your input signal, Plaxis will interpolate the values from the input file.

This may result in a different input signal than was intended. In general, it is recommended to choose *Δt*, *m* and *n* in such a way that δt is equal to the time interval used in the input signal.

In general, it is a good habit to pre-select a node on the lower boundary (where the prescribed displacement, used to model the earthquake signal, is defined). In Output, the user can plot a time-history signal for this node. The time-history signal found should be identical to the time-history signal used as input. This can be explicitly checked by comparing the numerical values from Output with the numerical values in the input file.