Estimation of accumulated fatigue damage in lattice support structures from operational vibrations

Abstract

Support structures of offshore wind turbines are prone to failure from fatigue damage. The design for fatigue requires accurate predictions of the environmental conditions and an adequate definition of the structural properties, valid for the entire design life-time. Estimates of the accumulated fatigue damage are, however, characterized by a large degree of uncertainty, stemming from the loading specifications and the numerical models used to predict the response. By employing measured data intelligently, the accumulated fatigue damage can be monitored throughout the structural life-time. This work presents a feasibility study towards the application of a joint input-state estimation algorithm for the response estimation of a lattice support structure. The feasibility is studied by first generating artificial measurement data with a full-order finite element model, while the strains at unmeasured locations are estimated with an erroneous reduced-order design model, after inclusion of measurement noise. It is shown that this model-based approach allows for the estimation of the response, despite significant errors in the design model. Particular attention is paid to the measurement locations, which should be within reach for maintenance.