Numerical Model Quality Assessment of Offshore Wind Turbine Supporting Structure Based on Experimental Data

Abstract

As a structure degrades some changes in its dynamical behavior can be observed, and inversely, observation and evaluation of these dynamical changes of the structure can provide information of structural state of the object. Testing of the real structure, besides of being costly, can cover only limited working states. It is particularly considerable in case of hardly accessible, and randomly/severely dynamically loaded offshore structures. As a testing instrument, numerical simulations are not limited as much, however a quality of answers depends significantly on an excellence of numerical model, and how in reality it reassembles the actual structure and the test results. One of the strategies is to correlate and update numerical models basing on the experimental data. Presented outcomes were obtained in frame of “AQUILO” project that aims to create a knowledge base, from which the investor will be able to decide on the best type of support structure for offshore wind farm specific location in Polish maritime areas. The examined object is laboratory tripod type support model (scaled) of the offshore wind turbine supporting structure, with appended flange on the one of the branches, allowing simulation of a fatigue cracking process. For the assessment and comparison with numerical model calculation results of the dynamical state of the structure the Experimental Modal Analysis (EMA) approach was selected. After several measuring campaigns a database of results including varying type of supporting condition, and crack opening stages, was obtained. The numerical model was constructed with use of Finite Element Method (FEM) approach. The quality of FE model was assessed using Modal Assurance Criterion (MAC) that compares both models modal vectors, that is modal deformation shapes. Also the differences in frequencies of modes was assessed and taken as quantification of an compatibility. The results (EMA) show importance of applying and modelling (FEM) of supporting condition. An additional senility analysis directs indicated best fit parameters for a start of FE optimization process. doi: 10.12783/SHM2015/349