Conference programme

Back to the programme printer.gif Print




Delegates are invited to meet and discuss with the poster presenters during the poster presentation sessions between 10:30-11:30 and 16:00-17:00 on Thursday, 19 November 2015.

Lead Session Chair:
Stephan Barth, ForWind - Center for Wind Energy Research, Germany
Jordi Olle Sanchez GE Wind, Spain
Co-authors:
OLLE SANCHEZ JORDI (1) F SANTIAGO MANEL (1) JOVE BELLOT JORDI (1)
(1) GE Wind, BARCELONA, Spain

Share this poster on:

Printer friendly version: printer.gif Print

Presenter's biography

Biographies are supplied directly by presenters at EWEA 2015 and are published here unedited

Mr. Jordi Ollé has been working in the wind industry for 3 years in collaboration with Alstom Wind. He is currently a R&D engineer in Alstom Wind in Transverse Technologies area (TT) at R&D department. His work is focused on develop new methodologies and tools for structural calculation, design advanced procedures models identification in experimental time series related to structural wind turbines components and correlation FEM models dynamically. Now he is leading a collaboration between Polytechnic University of Catalonia (UPC) and Alstom Wind to increase the know-how in statistical analysis and experimental data analysis.


Poster

Poster Download poster (12.15 MB)

Abstract

Multiaxial fatigue assessment including local component dynamics in submodeling

Introduction

Offshore Wind industry demands a reduction in the cost of energy (CoE) in order to be competitive to other sources of energy production. The trend is to develop bigger wind turbines to increase the ratio between produced energy to CAPital EXpenditures (CAPEX) and OPErational EXpenditures (OPEX). Implications on the structural design of components require lower stiffness-to-mass ratios. In consequence, inertial effects and local dynamics of components could become an important issue in next coming years.
Structural design is commonly performed using a two-steps approach. On the one hand, Loads Time Series (LTS) at component boundaries are provided by aero-hydro-servo-elastic global models, which are simplified geometrically but accounting for main global dynamics as well as main non-linarites. On the other hand, ratios between each unit load entry and the local stress at critical points inside the component are computed in Finite Element (FE) models. Such models represent the local detail but they are treated linear and quasi-static. Stress Time Series are obtained from a linear superposition of LTS times unitary input static FE analysis. In general, this approach is valid only when the spectral content of loads is in low frequencies compared the component local eigenfrequency. When the quasi-static hypothesis is not respected, local dynamical effects are potential stress contributors and alternative approaches are needed.

Approach

The proposed method uses a convolutive approach in order to account for component local dynamics. The convolution by means of Impulse Response (IR) functions is a technique broadly applied. Nevertheless, its present implementation accesses the information in differently. The modal basis of the component is used in order to project LTS over a new set of Modal Time Series (MTS). At the same time, eigenmodes are used to compute a new set of unit-input to stress response ratios on the component FE model. The advantage is two-fold: on the one hand this approach has an equivalent workflow to the quasi-static one. In practice it means using same tools and processes just slightly varying the information; on the other hand, the method allows using classical submodelling techniques to sequentially go from coarser component meshes to finer meshes representing local detail. Such techniques were normally a blocking point for dynamical calculations.

Main body of abstract

The present work deals with the mathematical implementation, validation and practical application to a real example. In a first step the theoretical consistency of the approach is proved and conditions for its application are found. The second part deals with the numerical validation of the method: a simple example in a beam is used to benchmark results against a transient time-domain calculation. In the same way, such test is extended to a simplified jacket structure. In both cases correlation with the reference results is excellent. Finally, in a third step a complete damage calculation is done on a jacket structure. Results are compared to the classical quasi-static approach in terms of stress error and damage quantification in order to get a general vision of the advantages of this new approach.

Conclusion

In the design of modern Wind Turbine Generators, using the mass more efficiently would mean to put under question the quasi-staticity of several components, allowing dynamic effects in fatigue design. The lack of efficient numerical approaches could become critical in the next coming years. The proposed methodology addresses the inclusion of the dynamics on the design validation of structural components. This is done preserving existent workflow, processes and tools and it allows classical submodelling techniques. The method is verified and implemented in a damage calculation of a real structure. Results are critically compared to other classical approaches.


Learning objectives
Show a feasible and efficient way to include local dynamic effects on multiaxial fatigue calculation. The method is demonstrated as an accurate method compared to time consuming unfeasible transient calculations. Also it is presented as an alternative to avoid the use of arbitrary dynamic amplification factors and prevent overconservative structural components.