Browse/search for people

Publication - Dr Alberto Pirrera

    Sensitivity of beam cross-sectional coupling stiffness terms to modelling simplifications and variations

    Citation

    Maes, V, Macquart, T, Weaver, P & Pirrera, A, 2019, ‘Sensitivity of beam cross-sectional coupling stiffness terms to modelling simplifications and variations’.

    Abstract

    Extensive work has been done on the analysis of wind-turbine blades using cross-sectional modellers. Comparison between varying techniques has shown significant discrepancies [1], which are not always justifiable based on the underlying theories. Some authors have taken this to mean that certain modellers are less accurate or robust in their implementation. Other authors have instead shown that at least some of the difference can be traced to small discrepancies in inputs as well as cross-sectional representation [2]. While the main diagonal terms of the compliance and stiffness matrices are generally less sensitive to small variations in geometry or material distributions, it can be shown that the off-diagonal coupling terms are more responsive to these differences. These coupling terms are becoming increasingly important in new designs and technologies, as researchers and businesses attempt to actively benefit from the associated aeroelastic tailoring of coupled blades.

    In this study, variations in the fidelity of geometrical details and other model simplifications are investigated for their effect on the cross-sectional compliance terms as predicted using BECAS as well as solid element models in ABAQUS. The cases investigated, such as the corners of webs or internal box sections shown in Figure 1, are often simplified or inconsistently modelled in the design and analysis of beam and blade like structures.

    The findings of this study show that more care and precision is needed, as the influence on the off-diagonal coupling terms are significant. Furthermore, it is shown that when inputs match well, the computed cross-sectional stiffness terms as computed by BECAS and using solid elements in ABAQUS also match well for a range of cross-sections. For more complex cross-sections, with fully populated compliance matrices, differences in the predicted values can still be found. This indicates fundamental differences between the two modelling approaches which warrants further investigation.

    [1] Chen et al., Wind Energy 13, 497 (2010)
    [2] Saravia et al., Renew. Energy 105, 55 (2017)

    Full details in the University publications repository