Abstract
The highly variable and complex flow at operational AWE heights (of ~200–500 m or more) impacts the loads on a given AWES design. This involves different flow regimes which can involve turbulence and other phenomena, producing transient events that may dominate AWES loads as well as operation. As reported previously, flow accelerations over length- and time-scales relevant for AWES systems are different than those for conventional wind turbines, and so are the appropriately filtered flow statistics. Following the work of Gaunaa presented here (at AWEC2024) using lifting-line simulation/code, and in conjunction with the parallel work of McWilliam using AWEbox (this is 1/3 of a joint investigation), we examine the statistics of aerodynamic forces on a rigid AWES device — and their relation to corresponding flow acceleration statistics. The analysis is done for a modelled device based on the AP2 system by AMPYX (nominal power 30 kW, wingspan 5.5 m). The python-based framework AWEbox[3] is used to solve for optimal flight paths over a range of wind speeds and shear (corresponding to AWES regimes). From observations, transient events at different scales, due to phenomena not always seen at lower heights, are embedded in the inflow; this can include gradients not represented in mesoscale-model or re-analysis data. The inflow timeseries to simulations includes all three velocity components, capturing effects specific to AWES (significant accelerations in multiple directions affect AWES, while HAWTs act as larger-scale filters, mostly responding to / affected by streamwise fluctuations). In this study we start pragmatically by neglecting the dynamic interplay between flow field, flight paths, and control system; this will be investigated in later studies. The sensitivity of forces or stresses on the rigid AWES to transient inflow (i.e., filtered velocity and acceleration component statistics) will be examined, for the first time; this also includes analysis relative to ‘typical’ turbulent flow at operational AWES heights. Timeseries of simulated forces will also be further fed to loads calculations (primarily wing flap-wise bending moment and tether tension), analysed and described by McWilliams et al. in an accompanying presentation.
Mark Kelly
Associate Professor
My research interest is in atmospheric boundary layer flows, flow statistics and modelling, and uncertainty quantification.
Michael McWilliam
Senior Scientist
My research interest is in Systems Engineering and Multi-disciplinary Design Optimization.