Ultimate Limit State Model Basis for Assessment of Offshore Wind Energy Converters.

This paper establishes the model basis regarding the ultimate limit state consisting of structural, loading, and probabilistic models of the support structure of offshore wind energy converters together with a sensitivity study. The model basis is part of a risk based assessment and monitoring frame...

Full description

Bibliographic Details
Main Authors: Thöns, Sebastian, Faber, M. H., Rücker, W.
Format: Article in Journal/Newspaper
Language:English
Published: 2012
Subjects:
Online Access:https://orbit.dtu.dk/en/publications/153c9fe1-477b-4fab-a6ce-5461e10f5054
Description
Summary:This paper establishes the model basis regarding the ultimate limit state consisting of structural, loading, and probabilistic models of the support structure of offshore wind energy converters together with a sensitivity study. The model basis is part of a risk based assessment and monitoring framework and will be applied for establishing the "as designed and constructed" reliability as prior information for the assessment and as a basis for designing a monitoring system. The model basis is derived considering the constitutive physical equations and the methodology of solving these which then in combina-tion with the ultimate limit state requirements leads to the specific constitutive relations. As a result finite element models based on shell elements incorporating a structural and a loading model are introduced and described in detail. Applying these models the ultimate capacity of the support structure and the tripod structure are determined with a geometrically and materially nonlinear finite element analysis. The observed failure mechanisms are the basis for the definition of the ultimate limit state responses. A probabilistic model accounting for the uncertainties involved is derived on the basis of literature review and measurement data from a prototype Multibrid M5000 support structure. In combination with the developed structural and loading models, sensitivity analyses in regard to the responses are peiformed to enhance the understanding and to refine the developed models. To this end, as the developed models necessitate substantial numerical efforts for the probabilistic response analysis predetermined designs of numerical experiments are applied for the calculation of the sensitivities using the Spearman rank correlation coefficient. With this quantification of the sensitivity of the random variables on the responses including nonlinearity the refinement of the model is performed on a quantitative basis.