Choice of Time Step |
 
|
Choice of Time Step |
|
Whether you opt for a fixed or variable time step, it is important to choose a time-step which picks up necessary detail in excitation and response. This applies to all time domain dynamic analyses, but it is particularly relevant to those which include significant non-linearities. As a general rule of thumb, it is recommended to use a time step of between 1/12th and 1/20th of the applied wave period in a regular wave analysis.
However, much smaller time steps may be required if the analysis includes any significant non-linearity, such as intermittent contact between a risers and a guide surface for example. Impact situations tend to cause structure response at high frequencies. This problem is exacerbated by the fact that, in general, the element length around the area of contact tends to be small, which can lead to increased structure response at these high frequencies.
If you are particularly concerned about the time step you have chosen, a useful exercise might be to run some sensitivity analyses, to examine sensitivity of results to the time step size. You would typically expect to see the solution converge towards a consistent set of results for all time steps below a reasonable threshold level.
Another general principle of finite element analysis is that in theory you should always attempt to match the temporal and spatial discretisations. This is a slightly more difficult concept to embody, and requires a considerable level of finite element experience. Basically it means that if you are using a relatively fine finite element mesh, then you should be using a relatively fine time step. Conversely, if you are using a relatively coarse mesh, then a relatively coarse time step will be suitable.