## Environment: Wind data |

OrcaFlex includes the effects of wind on:

- Vessels – see current and wind loads
- Lines – see hydrodynamic and aerodynamic loads
- 6D buoys – see lumped buoy added mass, damping and drag and spar buoy and towed fish drag
- 6D buoy wings – see wing type data

You may choose whether or not wind loads are included for vessels, lines, 6D buoys and 6D buoy wings.

Allows you to control how the wind is applied during the static analysis and the build-up period. Three options are available:

**Unramped**. The full, unramped wind velocity is applied during the static analysis and throughout the build-up period. If the wind type is time dependent (e.g. spectral, time-history, or full-field), then the wind velocity at the simulation start time is used in the static analysis.**From zero**. No wind is applied during the static analysis and the wind velocity is ramped to its full value during the build-up period.**From mean**. The mean wind velocity is applied during the static analysis and the wind velocity is ramped to its full value during the build-up period.

This data is unavailable when frequency domain dynamic analysis is enabled, and the mean speed is always used in the static analysis.

The air density is assumed to be constant and the same everywhere.

Used to calculate Reynolds number. The value here is fixed and cannot be edited.

Used to calculate the wind turbine unsteady aerodynamics.

Wind speed is assumed to be the same at all heights, unless a vertical wind speed profile is specified. To specify a vertical wind speed profile, you may define the wind speed variation with height above the mean water level (MWL) as a dimensionless multiplicative factor. To do so, you define a vertical variation factor variable data source. Negative factors may be used, allowing you to model reversing wind profiles.

A value of ~ means that there is no vertical variation.

If you are using the OCIMF model for wind load on vessels, the speed is expected to be that at an elevation of 10m (32.8 ft) above the mean water level (MWL). If you have the wind speed $v(h)$ at some other height h (in metres), then the wind speed $V(10)$ at 10m can be estimated using the formula $v(10) = v(h)\,(10/h)^{1/7}$.

Note: | The vertical wind variation profile data is not available, and is not applied, when full field wind is modelled using the TurbSim format. For full field wind of this type, the vertical variation in the wind velocity is specified directly in the external full field wind data. When full field wind is modelled using the Mann format, it is only applied to mean wind speed. |

Wind can be defined a number of different various ways, by setting the **wind type** to one of the following.

The wind is defined by specifying its **speed** and **direction**. The direction remains constant over time. The wind ramping dictates the wind speed during the static analysis and the build-up period.

The wind speed varies randomly over time, using a choice of either the NPD spectrum, API spectrum or the ESDU spectrum.

In these cases:

- The wind
**direction**remains constant over time. - The spectrum is defined by the
**ref. mean speed**(the 1-hour mean speed at an elevation of 10m (32.8 ft) above MWL) and the**elevation**above MWL at which the wind speed is to be calculated. From these, OrcaFlex calculates the mean speed at the given elevation and parameterises the spectrum which determines the statistical variation about that mean. If a value of ~ is specified for the elevation, it is taken to be that of the reference mean speed, i.e. 10m (32.8 ft) above MWL. The**view spectrum**button plots a graph of the resulting spectrum. - The
**min**and**max**frequencies bound the range considered by the spectral discretisation algorithm: only spectral energy between these frequencies will contribute to the wind velocity. The default values correspond to the range of frequencies for which the NPD spectrum is defined, as documented in ISO 19901-1:2005. To include*all*the energy in the spectrum, use values of 0.0 and infinity. If the max frequency is infinity, we use an approximation to integrate the tail of the spectrum out to infinity which requires that the min frequency must be set to 0.0. - The wind speed is modelled by the superposition of sinusoidal functions of time, their number given by
**number of components**, whose amplitudes and frequencies are chosen by OrcaFlex to match the spectral shape. OrcaFlex uses the same equal-energy approach to choosing the amplitudes and frequencies as for wave spectra discretisation. You should choose a number of components large enough to give a reasonable representation of the spectrum. - The wind ramping dictates the wind speed during the static analysis and the build-up period.
- The phases of the components are chosen using a pseudo-random number generator that generates phases which are uniformly distributed. The phases generated are repeatable – i.e. if you re-run a case with the same data then the same phases will be used – but you can choose to use different random phases by altering the
**seed**used by the random number generator. - The
**view wind components**button gives a report of the components that OrcaFlex has chosen. This will tell you the width of the frequency intervals, which can help you to judge whether the number of components is sufficient. - If the ESDU spectrum is selected, the absolute value of the site
**latitude**must be specified.

Note: | When frequency domain dynamic analysis is enabled, the mean wind speed is used during the static analysis, and the wind spectrum specifies the dynamic wind behaviour. |

A user-defined spectrum is given by a table of pairs of values of **frequency** $f$ and **S**, the spectral energy $S(f)$.

The given values of $f$ do not need to be equally-spaced. For intermediate values of $f$, OrcaFlex will obtain $S(f)$ by linear interpolation. S(f) is taken as zero for values of $f$ outside the range of the table. Your table should therefore include enough points to adequately define the shape of $S(f)$ (particularly where $S(f)$ has a wide range or high curvature) and should cover the full frequency range over which the spectrum has significant energy.

The above description of wind speed calculation for NPD, API and ESDU spectra applies equally to user defined spectra, with the following exceptions:

- The
**mean speed**is entered directly, rather than being calculated from the ref. mean speed and elevation. - The min and max frequencies are determined by the range of frequencies in the table.

Note: | When frequency domain dynamic analysis is enabled, the mean wind speed is used during the static analysis, and the wind spectrum specifies the dynamic wind behaviour. |

The wind is defined as the sum of a number of given sinusoidal components. For each component you give:

**Frequency**or**period**: you may specify either one of these – the other is automatically updated using the relationship period = 1 / frequency.**Amplitude**: the single amplitude of the component – that is, half the peak to trough height.**Phase lag**: the phase lag relative to the wind time origin.

The **randomise phases** button will generate a random phase value for each component, replacing all the existing data.

The wind speed variation with time is specified explicitly by time history. Linear interpolation is used to obtain the wind speed at intermediate times.

You must also provide **mean speed** and **mean direction**. The wind direction remains constant over time. The wind ramping dictates the wind speed during the static analysis and the build-up period.

Both the wind speed *and* direction variation with time are specified explicitly by time history. Linear interpolation is used to obtain the wind speed and direction at intermediate times.

You must also provide **mean speed** and **mean direction**. The wind ramping dictates the wind speed and direction during the static analysis and the build-up period.

The wind speed (at the shear reference origin), direction, linear horizontal shear, linear vertical shear and gust speed variation with time are specified explicitly by time history. Linear interpolation is used to obtain the values at intermediate times.

A **shear reference origin** and a **shear reference length** must always be specified. The shear reference origin, $(X\urm{ref}, Y\urm{ref}, Z\urm{ref})$, is the global position at which the time varying wind speed, $U\urm{ref}(t)$, is unmodified by shear (although it is still affected by the vertical variation factor). The shear reference length, $l\urm{ref}$, is then used, along with the time varying linear horizontal shear, $S\urm{H}(t)$, and linear vertical shear, $S\urm{V}(t)$, to calculate the spatial variation of the wind about the shear reference origin. Vertical shear is in the global $Z$ direction. Horizontal shear is perpendicular to both the wind direction and the vertical. For positive horizontal shear, if the wind direction is zero, i.e. the wind is propagating along the global X direction, the wind speed will increase in the global Y direction. The time varying gust, $U\urm{gust}(t)$, acts independently of any shear and vertical variation factor, and is added to the wind speed at all points in space. The wind speed, in the time varying direction, $\theta(t)$, at the global position $\vec{p} = (X, Y, Z)$, and time $t$, is calculated via
\begin{equation}
u(\vec{p}, t) = U\urm{ref}(t) \left(F\urm{V}(Z) + S\urm{V}(t)\frac{Z-Z\urm{ref}}{l\urm{ref}} + S\urm{H}(t)\frac{\left(Y-Y\urm{ref}\right)\cos{\theta(t)}-\left(X-X\urm{ref}\right)\sin{\theta(t)}}{l\urm{ref}} \right) + U\urm{gust}(t)
\end{equation}
where $F\urm{V}(Z)$ is the vertical variation factor.

You must also provide **mean speed** and **mean direction**. The wind ramping dictates the wind speed and direction during the static analysis and the build-up period. The shear and gust are assumed to be zero in the statics calculation.

Full field wind allows for variation of wind velocity in both space and time, with data specified in external files. The coordinate system used in the files is right-handed, with $x$ horizontal in the direction of propagation, $y$ horizontal and normal to $x$, and $z$ vertically upwards.

To use full field wind, you must always define the following:

- The wind
**direction**and**origin**, which determine how the file's coordinate system is mapped on to the OrcaFlex coordinate system. A value of ~ for the $Z$ wind origin can be used to specify that the vertical origin is at the mean water level. - The wind time origin. This is the simulation time corresponding to a time of zero in the external file.

Two binary **full field wind formats** are supported: TurbSim .bts files; and Mann turbulence generator .bin files.

When using the **TurbSim** format, the **name** of a single .bts file must be specified. You can give either its full path or a relative path. Clicking **file header** allows you to view the information contained in the .bts file header. The TurbSim .bts file contains time series of 3D wind velocity, $\vec{V}\urm{g}(y,z,t)$, at points on an evenly spaced grid in the vertical $yz$ plane. Optionally, the file may also contain time series of 3D wind velocity, $\vec{V}\urm{t}(z,t)$, at tower points in a single line below the grid.

When using the **Mann** format, you must define the following:

- The
**names**of three .bin files, corresponding to the u, v and w components of the wind field. Respectively, u, v and w correspond to the wind velocity component in each of the x, y and z directions defined by the file's coordinate system. You can give either their full paths or relative paths. - The
**mean speed**in the direction of propagation, to which the vertical variation factor is applied, before being added to the file's grid velocity. This must be greater than zero. - The grid size, defined by specifying the number of grid points in each direction, ($N_x, N_y, N_z$), and the spacing between them, ($\textrm{d}x, \textrm{d}y, \textrm{d}z$). The grid size must be consistent with the data contained in the named files.

Each Mann .bin file contains a 3D spatial grid of a single wind velocity component, $V_{\textrm{g},i}(x,y,z)$, such that $\vec{V}_\textrm{g} \equiv (V_{\textrm{g},1}, V_{\textrm{g},2}, V_{\textrm{g},3})^\textrm{T}$. The grid's origin is at the centre of its $yz$ plane. Optionally, each velocity component can be scaled before it is used in OrcaFlex. This is done by specifying the component's **target standard deviation**. A value of ~ means that there is no scaling and the component is to be used unmodified. Otherwise, OrcaFlex first estimates the file's standard deviation, along the direction of propagation, at a data point as close to the centre of the grid's $yz$ plane as possible. OrcaFlex then determines the factor needed to scale the standard deviation estimate to the target value, and applies it to the whole grid before it is used.

For both formats, OrcaFlex uses Taylor's frozen turbulence hypothesis. This uses the mean wind speed (as recorded in the .bts file for TurbSim, or given as data for the Mann format), to map between the 3D grid contained in the files (i.e. $\vec{V}\urm{g}(y,z,t)$ for TurbSim, or $\vec{V}\urm{g}(x,y,z)$ for the Mann format) and $\vec{V}\urm{g}(x,y,z,t)$ as required by OrcaFlex; or, in the case of the TurbSim tower, between $\vec{V}_t(z,t)$ and $\vec{V}_t(x,z,t)$.

To interpolate in the grid, OrcaFlex uses barycentric interpolation. For points outside the grid, OrcaFlex clips to the edge of the grid, along each primary axis. For example, consider a .bts file with no tower points, and with a grid defined at $y_1, y_2, \ldots, y_{N_y}$ and $z_1, z_2, \ldots, z_{N_z}$. For values of $y < y_1$ or $y > y_{N_y}$, OrcaFlex clips $y$ to $y_1$ or $y_{N_y}$ respectively. Similarly, for values of $z < z_1$ or $z > z_{N_z}$, OrcaFlex clips $z$ to $z_1$ or $z_{N_z}$ respectively.

The TurbSim .bts file format supports periodic time histories. If the file is periodic, as recorded in the file header, OrcaFlex will interpret the data accordingly. For non-periodic files, if extrapolation in time is required it is performed by clipping to the defined range. The Mann file format is always assumed to be periodic in the direction of propagation.

To help visualise the full field wind, a wire frame 3D box and arrows representing the vector field can be drawn.

The wind ramping dictates how the wind is applied in the static analysis and during the build-up period.