Skip to content

Latest commit

 

History

History
180 lines (97 loc) · 11.6 KB

ReleaseNotes.md

File metadata and controls

180 lines (97 loc) · 11.6 KB
Raw

IEA Wind 15-MW Release Notes

v 1.1.6

  • Addition of HAWC2 monopile model

  • Alignment of tower and monopile discretizations and properties across YAML, HAWC2, and OpenFAST files

  • Addition of SolidWorks CAD model for UMaine VolturnUS-S semisubmersible

v 1.1.5

  • Update OpenFAST input files for v3.5.1 release API changes

  • Floating tower damping reverted to 1.0% to resolve numerical instabilities. There is a good deal of uncertainty about the true value here as it is normally determined experimentally. Comments to this effect have been added to the Wiki entry.

  • Update the second-order potential-flow wave excitation of the VolturnUS-S platform. PR #137.

v 1.1.4

  • Update OpenFAST input files for v3.4.1 release API changes

  • Update ROSCO libraries and inputs to v2.7

  • Blade elastic properties consistently originate from SONATA/ANBA4

  • Cleanup of stale files and README comments

v 1.1.3

  • Update OpenFAST input files for v3.3 release API changes

  • Update ROSCO to v2.6

  • Provide both WAMIT input and output files, with a bug fix regarding platform orientation

  • Adjust OpenFAST ElastoDyn NacYIner and PtfmYIner according to monopile simulation best practices

  • Additional references to community design contributions and permutations in the README

  • Remove deprecated BModes files

  • Fix minor errors in the tabulated documentation data

v 1.1.2

  • Update ROSCO to v2.5

v 1.1.1

  • Fix OpenFAST floating tower mode shape coefficients to avoid instability

  • Fix incorrect OpenFAST VolturnUS-S PtfmCMzt value to WISDEM-generated value and closer to original value

v 1.1

This update to the IEA Wind 15-MW Reference Wind Turbine attempts to address some of the modeling issues that have been pointed out by users. Larger redesign requests, such as a smaller diameter monopile, or higher specific power for the floating turbine, are not included as that would require a more significant allocation of resources and will likely be left to the broader community to generate design alternatives. Please see the Wiki FAQ for some extended answers to those topics.

Major Changes

  • Many of the modeling tools have undergone changes and updates, so the files here have been updated to keep pace with API changes.

    • WISDEM has undergone perhaps the most extensive overhaul. It now uses the YAML-based ontology files from the WindIO project as the chief input, and is also more tightly integrated with OpenFAST and ROSCO. More detailed changes described below.

    • OpenFAST has seen a couple of releases since the 15-MW RWT input files were created. Additionally, the original input deck also had a number of errors and inconsistencies. The integrations with WISDEM and, by extension, the WindIO ontology files have helped to rectify these problems. Finally, a set of input files that works with OLAF, the OpenFAST vortex particle aerodynamics code has been included.

    • ROSCO, which provides the controller, was a new tool when the 15-MW RWT was first released and has since seen numerous bug fixes and new features. More details described below and at the ROSCO repository.

  • The WindIO ontology YAML-files have been expanded to include descriptions of the entire turbine. This follows the progress made in the parallel IEA Task 37 WindIO project. Note that the YAML descriptions for some of these components will likely still evolve, especially the nacelle components, as settling on a communal parameterization takes time and numerous iterations.

  • Draft version of the ontology for the UMaine Volturn-S semisubmersible design variant is now included in the WT_Ontology directory. This should be considered a work in-progress as there is still extensive work to do to verify how well the description matches the original intent of the design report.

  • There was an inconsistency in the stated hub diameter in the original report, YAML-definition, and WISDEM files (7.94m) versus the hub diameter in the OpenFAST and HAWC2 files (3.0m). This was due to a late design change to better accommodate the blade root chord/diameter size that did not fully propagate to all input files. With the hub diameter at 7.94m and the blade unchanged at 117.15m in length (arc length measurement that includes the prebend), the precise rotor diameter is 242.24m. This is called out in Issue #51 and the Wiki FAQ.

  • The nacelle component mass properties and dimensions were approximate in the report and then inconsistently captured between the various model input decks. Furthermore Table 5-1 in the report had values that could do be replicated and did not add up (called out in Issue #30. With the new capabilities in WISDEM's DrivetrainSE model, this has been rectified in the nacelle mass properties tab in the IEA-15-240-RWT_tabular.xlsx file in the Documentation folder.

  • The floating tower design demonstrated resonance excitation from the 3P rotor mode in both OpenFAST and HAWC2, suggesting that there was an error in the original design optimization script. The floating tower has been redesigned to be much stiffer, resulting in a sharp increase in tower mass around a similar tower center of gravity. This was also called out in the Wiki FAQ.

  • Bug fixes and other improvements in composite material representation in the WISDEM model has resulted in a slight growth in blade mass compared to the original report (68t vs 65t). This was also called out in the Wiki FAQ. Nevertheless, the composite material properties, especially for carbon fiber reinforced carbon (CFRC) composites, are known to be out of date and inconsistent with modern pultrusion manufacturing methods. A future release will likely update these properties and re-optimize the spar cap thickness along the blade.

  • Trailing edge reinforcement was mistakenly added too close to root and tip of blade in the first release, and was not always consistent with the physical space constraints along the span. Due to different approaches to meshing the trailing edge, this also created some inconsistent stiffness distributions depending on model inputs. This has been corrected via this Pull Request.

Detailed OpenFAST changes

For the monopile variant:

  • InflowWind:

    • RefLength corrected to rotor diameter
    • API changes to be consistent with OpenFAST v3.0

  • AeroDyn:

    • Blade mass and stiffness properties have been updated per the updates to the material properties.
    • The blade mode shapes for ElastoDyn were incorrect in multiple ways. In the initial release, the flap and edge modes were both the first bending mode (as captured in Issue #37. Furthermore, there was a slight error in the polynomial fitting routine that led to some non-physical trends in ElastoDyn in flexible blades. Both of these issues have been corrected.
    • API changes to be consistent with OpenFAST v3.0

  • ElastoDyn:

    • Zero-ed out initial platform heave, PtfmHeave.
    • Made small changes to (TipRad, HubRad, Overhang, Twr2Shft, TowerHt) to match ontology values
    • NacCMxn/yn/zn updated based on bottom-up WISDEM DrivetrainSE calculation and tracking
    • Nacelle component masses and MofI calculated through WISDEM DrivetrainSE. Values for HubMass, HubIner, GenIner, NacMass, NacYIner, YawBrMass, DTTorSpr, and DTTorDmp all updated accordingly.
    • PtfmYIner zeroed out because transition piece is accounted for in SubDyn
    • Tower modes shape coefficients were updated with corrections to the polynomial fit routines

  • HydroDyn

    • Made small changes to (WaveHs, WaveTp) to match ontology values
    • Monopile geometry specification is the same as SubDyn and the ontology file, with more accurate thickness profile
    • Activated ExctnMod by default per Issue #52

  • SubDyn:

    • Transition piece added as a concentrated mass at top of monopile

  • ServoDyn:

    • Small change to GenEff to match bottom-up estimate (from 96.55 to 95.756)
    • Default GenModel switched to 1 (simple)
    • Zeroed HSS brake data since that has not been specified in the design
    • YawSpr set to torsional stiffness of tower and YawDamp set for 1% damping ratio for first tower torsional mode
    • YawManRat changed from 2 deg/s to 0.25 deg/s due to size of RNA
    • Turned off DLL_DT and DLL_Ramp

  • MoorDyn:

    • API changes to be consistent with OpenFAST v3.0

For the UMaine Volturn-S floating variant:

  • All of the changes associated with the monopile variant
  • RotSpeed set to 7.55, the rated speed
  • Reset initial platform heave, PtfmHeave, to keep the same MSL point on the platform with the nacelle mass updates
  • Redesigned the tower to be significantly stiffer giving a first fore-aft/side-side frequency around 0.49Hz.

Also added input files that invoke OLAF, the OpenFAST vortex particle method for aerodynamics (monopile variant only)

Detailed HAWC2 changes

TBD

Detailed WISDEM changes

Three WISDEM scripts remain:

  • run_model.py : WISDEM analysis (no design optimization) of the IEA Wind 15-MW RWT. Uses the WindIO ontology file directly as input, along with modeling_options.yaml and analysis_options.yaml. Script also includes automated plotting of blade geometry, rotor performance, and tower geometry. The tabular data Excel sheet is also generated.

  • optimize_monopile_tower.py : Script that can be used to optimize the tower and monopile diameter and thickness schedule. Uses the WindIO ontology file directly as input, along with modeling_options_monopile.yaml and analysis_options_monopile.yaml.

  • optimize_floating_tower.py : Script that can be used to optimize the tower diameter and thickness schedule on the Volturn-S floating platform. Uses the Volturn-S WindIO ontology file directly as input, along with modeling_options_floating_tower.yaml and analysis_options_floating_tower.yaml.

  • optimize_generator.py : Script that can be used to optimize the permanent magnet synchronous generator.

For more information about WISDEM and how these scripts can be run, please see the WISDEM documentation and repository.

Detailed ROSCO changes

v 1.0

  • Jan 2021: Officially tagged the v1.0 release, although there are still some outstanding issues

  • Apr-Nov, 2020: Series of minor fixes in model input files (see closed pull requests for details)

  • March 30, 2020: Finalization of OpenFAST and HAWC2 models in conjuction with release of report and media pieces