Dynamic Calculation Parameters

This editor contains all parameters available to configure the dynamic calculation of a RIFLEX task.

The main sections of Dynamic calculation parameters are;

  • Irregular analysis parameters related to irregular wave analysis

  • Regular analysis parameters related to regular wave analysis

  • Procedure parameters related to time domain integration, damping and dynamic forces

  • Loading, change of static loads and various dynamic loads including boundary change

  • Storage of results

  • HLA settings

Special care should be used when changing integration and damping parameters.
It is recommended to choose set to simulator default under Procedure to introduce some numerical damping for long analyses.

The binary format must be selected for storage if the results are to be used in the SIMA post processor.

Note that the actual choice of regular/irregular analysis is set in the condition.

1. Main Simulation Parameters

The following parameters are available:

  • Simulation length

  • Simulation time step

The simulation time step should be chosen such that important dynamic effects are captured in the structural analysis. This time step should be chosen based on the loading rates, finite element length, dynamic response, and iteration procedure. Simulations considering contact problems generally require a short time step.

2. Time Series Generation Parameters

The following parameters are available:

  • Wave seed number

  • Use stochastic amplitudes

  • Requested time series length

  • Time increment

Time series of waves, wind, and/or first or second order wave loads (or linear support vessel motions) are required during the dynamic simulation. These time series should be generated such that they are at least as long as the full dynamic simulation and correctly represent the desired frequencies. That is, the time series length should be greater than or equal to the simulation length (to avoid repetition), and the time increment for time series generation should be appropriate for the given process. The time increment for time series generation should also be an integer multiple of the simulation time step.

The wave seed is used to generate random phase angles for each wave frequency component, giving different realizations of the selected wave spectrum for different seeds.

If stochastic amplitudes are not selected, deterministic amplitudes are used (this is the default option). All realizations will have the same spectrum with sampled or interpolated values from the specified wave spectrum.

If stochastic amplitudes are selected, the amplitude of each wave component will have a scatter around the values given by the wave spectrum. The realizations will have different spectra and the significant wave height will vary between realizations. Multiple, long wave realizations should therefore be simulated if stochastic waves amplitudes are selected.

As shown below, the time series (wave/wind/load/motion) are generated by discretizing the variance spectrum into a finite number of harmonic components with uniformly distributed phases using FFT. The phase angles are generated using a pseudo-random number generator with the seed given as user input. Choosing a different seed will give a different time series realization which in turn will yield different simulation results when wave loads or support vessel motions are included in the simulation.

image

The time increment, frequency increment and maximum duration (without repetition) are interrelated when the addition of the harmonic components is performed by FFT:

\(\Delta \omega = \frac{2\pi}{N_t \Delta t} = \frac{2\pi}{T_{rep}}\)

where \(\Delta \omega\) is the frequency increment, \(N_t\) is the number of time steps (in practice, this will be increased such that \(N_t = 2^N\)), \(\Delta t\) is the time increment, and \(T_{rep}\) is the length of the generated time series.

3. Irregular Wave and Motion Timeseries Parameters

This section includes options for the generated wave and motion timeseries:

  • Simulation start time: Specifies at which point in time on the pregenerated timeseries the simulation should start

  • Irregular wave: Wave forces may be included on RIFLEX elements

  • Irregular and low-frequency support vessel motions: Options for prescribed support vessel motions

  • Add wave time series from file: Option for reading a wave time series from file. If this is selected, the wave spectrum given in the environment input will be ignored.

  • Motion scaling: Enable scaling of motions for support vessels

If wave forces are included on RIFLEX elements (Wave forces present) the Wave kinematics specification section will be available.

4. Wave Kinematics Specification

SIMA allows the user to control the application of wave kinematics at nodes. The nodes for calculation of wave kinematics should be chosen such that important wave loads are captured. Reducing the number of nodes for kinematics calculation improves the simulation efficiency.

The main input is:

  • Default procedure for wave kinematics calculation on/off

  • Node step

  • Z Lower

  • Z Upper

  • Diffracted wave kinematics at nodes

  • Specification of node-step for individual lines

Wave kinematics may be pre-generated using FFT or calculated during the simulation as a sum of cosine terms. Pre-generation is much more efficient than calculating kinematics during the dynamic simulation.

Kinematics can be pre-generated at the static positions of the selected kinematics nodes.
Alternatively, kinematics can be calculated at the dynamic position of the selected kinematics nodes by summing cosine terms at each generation time step.

The kinematics at a generation time step are calculated using the dynamic position at the previous generation time step. Kinematics at simulation time steps (between the generated time steps) are found using linear interpolation. The generation time step should be small enough that the change in dynamic position between generation steps is negligible and that the highest frequency wave components with significant contributions are well represented. Om the other hand, a short generation time step will lead to frequent calculation of kinematics, which will make a time-consuming simulation.

Selecting kinematics calculated during the simulation for a linear time domain simulation with Wilson’s integration method will result in the calculation of kinematics at two generation time steps each time a new generation time step is reached.

The option Kinematics at static position calculated during simulation should give the same results as Kinematics at static position, but take several times longer. It is primarily implemented for for testing.

The option Kinematics at fixed static position (linearized) results in the same kinematics as the default option Kinematics at static position. The difference is that in Kinematics at fixed static position (linearized), a node remains wet or dry depending on its static position. Thus, a node that has a static position 10 cm below the surface will continue to have wave loads even if it is lifted well above the surface in a dynamic simulation. The option may be useful when investigating differences between nonlinear and linear analysis.

The possible combinations are described below.

4.1. Default procedure on, default Z lower/upper, Nodestep = 1

This is the recommended default option in SIMA and will apply wave kinematics at all nodes between:

  • The water depth entered under Location (lower limit)

  • Four times the square of the standard deviation of the total wave elevation (upper limit)

4.2. Default procedure on, Z Lower and Upper given by user

If Z Lower and Z Upper is set to editable, the user may manually limit the nodes where wave kinematics will be applied in the dynamic analysis. Wave kinematics will not be applied to nodes outside this range. This range will be applied to all lines in the system.

4.3. Default procedure on, Nodestep != 1

If Nodestep is set to an integer value different from 1, the value will be used as nodestep for calculating wave kinematics. Linear interpolation will be used for intermediate nodes. A nodestep of 2 will for example result in calculation of kinematics at every other node and linear interpolation for intermediate nodes.

4.4. Default procedure on, Nodestep < 1

A negative integer value may be given for nodestep. The distance between Z upper/lower is then divided into 4 equal intervals and nodestep is increased gradually giving the following nodesteps:

  • |nodestep| in the upper interval

  • 2 * |nodestep| in the second interval

  • 4 * |nodestep| in the third interval

  • 8 * |nodestep| in the fourth interval

This option is intended to reduce computation time when generating timeseries of wave kinematics. The option should be used with care as the resulting response is dependent on element length, water depth and system type. Nodestep < –1 is therefore not recommended.

4.5. Default procedure on, diffracted wave kinematics at selected nodes

If a transfer function file for wave kinematics is included, the default kinematics will be replaced for the specified nodes.

  • If one node is specified, diffracted kinematics will be applied at this node.

  • If several nodes are specified, diffracted kinematics will be calculated at these nodes and linear interpolation of diffracted kinematics will be applied to intermediate nodes.

4.6. Default procedure on, Nodestep specified for selected lines

Nodestep may be specified for individual lines. This specification will replace the default specification for the selected line.

4.7. Default procedure off, Nodestep specified for selected lines

Wave kinematics will only be applied to the lines that are specified with a nodestep.

Z Upper, Z Lower, and nodestep given under default procedure are now dummy values.

4.8. Default procedure off, Diffracted wave kinematics at selected nodes

Diffracted wave kinematics will only be applied to the nodes that are specified.\\

  • If one node is specified, diffracted kinematics will be applied at this node.

  • If several nodes are specified, diffracted kinematics will be calculated at these nodes and linear interpolation of diffracted kinematics will be applied to intermediate nodes.

4.9. Wave kinematics specified for selected nodes

Wave kinematics may be read from file and applied at specified nodes. The wave kinematics file must contain wave elevation, velocities in the global X, Y and Z directions and accelerations in the global X, Y and Z directions for each position which will be used.

  • If one node is specified, the input kinematics will be applied at this node.

  • If several nodes are specified, the input kinematics will be applied at these nodes and linear interpolation of kinematics will be applied to intermediate nodes.

5. Second Order Wave Kinematics

Second-order Stokes wave theory for finite water depth may be used to calculate fluid velocities and accelerations below the water surface if the second order wave kinematics option is chosen. The first order wave elevation and mean water depth are input to the calculation. A second order, non-linear contribution to the wave elevation is calculated and added to the first order elevation. The second order contribution grows with increasing wave steepness, leading to a flattening of troughs and sharpening of crests. Fluid velocities and accelerations are calculated from the velocity potential, up to and including second order. Currently, second order wave kinematics are only available for long-crested sea-states.

6. Storage

To be able to see the results from within SIMA, the user must choose the Binary Format storage.

6.1. Visualization response storage

The user has two options when it comes to storing responses for visualization. The default option is SimVis, which will result in an ldat file. This is the format SIMA use for showing static and dynamic results in the 3D graphics. The VTF option will result in a vtf file. This may then be visualized in programs such as Ceetron GLView and SESAM Xtract.

6.2. Transformation Matrix and Angle Calculations

Time-series of angles may be calculated based on element transformation matrices. The transformation matrix may be stored for elements in the dynamic analysis and further post-processed to calculate the angle between local x-axis of two elements.

The transformation matrix \([\Lambda\)] contains direction cosines between axes of the global coordinate system \(x,y,z\) and the local coordinate system \(x^{’},y^{’},z^{’}\). A vector \(V\) can be expressed in terns of components \(u,v,w\) in system \(x,y,z\) or in terms of components \( u^{’},v^{’},w^{’}\) in system \( x^{’},y^{’},z^{’}\).

\[\begin{Bmatrix} u^{’}\\ v^{’}\\ w^{’} \end{Bmatrix} = [\Lambda] \begin{Bmatrix} u\\ v\\ w \end{Bmatrix}\]

\([\Lambda\)] is an orthogonal matrix, hence the inverse transform is

\[\begin{Bmatrix} u\\ v\\ w \end{Bmatrix} = [\Lambda]^{T} \begin{Bmatrix} u^{’}\\ v^{’}\\ w^{’} \end{Bmatrix}\]

The cosine of an angle between a local and global axis is found directly from \([\Lambda\)].For examle, the angle between local x-axis and global z-axis (for each timestep) \(\phi_{xz}\) is

\[\phi_{xz} = acos(\Lambda(1,3))\]

The angle between local x-axis and global y-axis

\[\phi_{xy} = acos(\Lambda(1,2))\]

The angle between local x-axis and global x-axis

\[\phi_{xx} = acos(\Lambda(1,1))\]

The angle between the local x-axis of two elements is

\[\phi_{xx} = acos(V_{x1} \cdot V_{x2} )\]

where

\[V_{x1}= [\Lambda_{1}]^{T} \begin{Bmatrix} 1\\ 0\\ 0 \end{Bmatrix}\]
\[V_{x2}= [\Lambda_{2}]^{T} \begin{Bmatrix} 1\\ 0\\ 0 \end{Bmatrix}\]

Note that the post-processor will return the smallest positive angle between the elements/axes in degrees.

Currently, angle between two elements (relative element angles) may be requested in storage parameters.