Version 3.2.0 Release

This release fixes some bugs with dtocean-wec that occurred if the wrong type of mesh file was loaded (i.e. a .stl file). This involved a complete overhaul of the dtocean_wec.submodule.utils.mesh_util module which has now been renamed as dtocean_wec.submodule.utils.mesh.

What's Changed

• Catch some bugs when bad mesh files are used in dtocean-wec by @H0R5E in #14

Full Changelog: v3.1.1...v3.2.0

Version 3.1.1 Patch Release

This patch release removes a debugging print statement added in #13.

Full Changelog: v3.1.0...v3.1.1

Version 3.1.0 Release

This release changes the way in which the non-code data files required for the package are discovered. They are now found from the dtocean-data conda package or the Windows installer available from the dtocean-data repository.

What's Changed

• Move to using the dtocean-data package for non-code data by @H0R5E in #13

Full Changelog: v3.0.0...v3.1.0

Version 3.0.0 Release

This major release adds a number of improvements to tidal turbine modelling, many of which are detailed in the following article:

Topper MBR, Olson SS, Roberts JD. Techno-Economic Modelling of Tidal Energy Converter Arrays in the Tacoma Narrows.
Journal of Marine Science and Engineering. 2020; 8(9):646. https://doi.org/10.3390/jmse8090646

Detailed changes are as follows:

Added

• Added loop to rotor interaction solver. This produces a more accurate result than the single iteration used previously.
• Added check for additional grid orientations to find the maximum number of devices that can be placed during an optimisation. One additional device is added in each direction (in combination) and then assessed.

Changed

• Improved log messages
• Now uses the dominant wake model for superposition of rotor wake velocity deficits. The same dominant wake is chosen for the turbulence kinetic energy factor calculation.
• Replaced term "induction" to avoid confusion. Prefer "velocity" or "velocity coefficient".
• Replaced Mannings number input with beta and alpha coefficients for use with Soulsby type velocity profile calculation.
• Improved performance of rotor CFD data reading.
• Renamed deg360_to_radpi function as bearing_to_radians.
• Improved streamlines plot.

Fixed

• Fixed turbine angle of attack calculation.
• Fixed minimum distance between turbines check.
• Fixed rated power per rotor for MCT style device.
• Fixed rotor streamline calculation travelling in wrong direction.
• Fix bug in positioning of rotors for MCT style device.
• Fixed expected device layout orientation in optimiser by making the inter-column spacing refer to the x-direction and inter-row to the y. Array rotation is then the angle of the oncoming flow minus 90 degrees.
• Fixed optimiser start point estimator increasing scale when it should be reducing it
• Fix incorrect sorting of outputs for rotors by making names equal length
• Fix bug which forced all rotors to share the same input data

Version 2.0.0 Release

This release coincides with the release of DTOcean 2.0.

A new hydrodynamic data package is attached to this release

The following changes were made:

Added

• Added change log.
• Added continuous integration configuration files.
• Added output of calculated power matrix for the isolated device.
• Instantaneous device power may no longer exceed the rated power.

Changed

• Changed wave modules mean power output per device to return actual power
  generated rather than the sum of all powers across the power matrix.
• Optimised EnergyProduction function by reducing number of calls to
  block_diag. This has provided a one third reduction in run times.
• Reduce memory consumption of wave calculations by using single precision
  complex numpy arrays. This is necessary for solving large OEC farms with 8MB
  RAM.
• A alternative configuration file is now used for the location of the
  hydrodynamic data files if the module is bundled into the installer or if
  installed from source.

Removed

• Removed pin of numpy at version 1.10.1 and updated bootstrap command in
  setup.py, which now requires MinGW to be installed separately. Note that PyQt
  and matplotlib must still be pinned due to incompatibility of later versions
  with Python 2.
• Removed installer code and data provided in the DTOcean data package.

Fixed

• Numerous PEP8 fixes.
• Tidal module velocity profile generator switched to Manning's formulation as
  was incorrectly set to the Soulsby type.
• Fixed bugs in the array layout optimiser code.
• Fixed bug in calculation of device depths for wave module approximation test.
• Fixed depreciation warning when sending arguments to setup.py test.
• Refactored distance_from_streamline to improve readability and correct issue
  with streamlines travelling upstream rather than downstream.
• Fixed issues with using non-rectangular domain in the tidal module.
• Fixed issues determining depth excluded zones with non-rectangular domains.
• NaNs are now set to zero in interp_at_point and edge cases are better
  handled.
• Fixed confusing variable names for inputs of wave energy period and peak
  period for wave energy calculations.
• Fixed bug where the angle of attack for yawed tidal turbines was being
  incorrectly calculated.
• Fixed tidal current streamline plotting (when debug flag is True).
• Fixed bug in dtocean_waves' Directional class which reduced energy output from
  wave energy calculations.
• Fixed bug in conversion of compass bearings to trig angles.
• Fixed bug in array main direction setting in tidal energy calculations.

v1.0.0: Update README.md

Changed link to polite source code