In order to create a high-fidelity multi-physics, mechatronic and multi-scale simulation framework for wind turbines, different models and simulations software have been developed. An interface between the different models and simulations software have been developed. The inputs required by each model are computed from the wind turbine characteristics and from the outputs of the models located further upstream in the computational chain.. The interface between different software has been designed through a a survey of all the users of the different model outputs in the consortium.Their data requirements have been collected. A format and data structure of various inputs and outputs have been defined.
When it comes to modelling preliminary models for atmospheric, park, fluid-structure interaction, noise and 3D fatigue delamination have been developed.
WRF-LES meso-scale atmospheric model has been properly setup and benchmarking of these models have been verified using available data.
Park model based on the actuator surface (ASM) concept for designing wind parks have been implemented in OpenFoam architecture. The models have been validated with Lillgrund wind park data.
CFD software STAR-CCM+ has been utilized for understanding flow around the turbine. To understand the fluid-structure interaction, two methods Samcef Mecano-vortex and Samcef mecano-CFD have been developed. An interface for co-simulation between Samcef Mecano and STAR-CCM+ have been implemented and tested on models created from the NREL defined 5MW turbine.
In order to model the sound field in the near-field region, two approaches have been evaluated. The first one, developed by IVKDF, is based on Amiet’s theory. The second one, developed by SISW, is based on the combination of a stochastic method to compute sound sources with a finite element method to propagate sound waves. The Amiet-based strategy was found to be significantly less computationally expensive and was selected for the noise prediction of wind turbines. In order to model the sound field in the far-field region accounting for weather effects, the near-field noise method has been coupled with a far-field noise method based on ray-tracing. The applicability of the noise prediction strategy has been demonstrated for an isolated Siemens 2.3 MW wind turbine.
A novel 3D fatigue delamination method based on cohesive zone modelling has been developed. The method incorporates novel features such as a growth driving direction criterion, 3D J-integral evaluation in cohesive interfaces, as well as a cycle jump fatigue method based on cohesive elements. The method has been validated and implemented in Samcef. For intra-laminar damage, a 4-point bending test configuration setup have been developed for measuring intra-laminar stresses and strains. An intra-laminar damage model and identification tool have been developed and implemented in Samcef exploiting the results from the 4-point bending test campaign for fatigue material characterization. The two damage models (inter- and intra-laminar) have been coupled with a damage interaction criterion and a joint solution strategy. The work related to the effect on load history effects during fatigue damage is ongoing but preliminary results indicate that load history effects has an important effect on delamination damage development.
All models and software tools have been installed and tested on the Fraunhofer ITWM computing infrastructure. Each software is encapsulated within an individual and exchangeable software container. In addition, two model order reduction approaches have been tested.
A literature review on the social and environmental factors affecting wind energy development was carried out. From this review, three areas of public concern with respect to wind energy emerged: concerns about wind energy technologies, concerns about landscapes in which wind energy is to be implemented and concerns about the process of wind energy development, especially the degree of public participation. A methodology to collect stakeholder data, based on insights from the literature data have been proposed.