Speaker
Description
A dynamical model is proposed regarding the forces developed on a bluff body, typically a cylinder, placed in a constant uniform incompressible fluid flow. The model concerns two-degree-of-freedom forced motion experiments under high in-line motion amplitudes and it combines experimental data analysis, neural networks and parametric modeling approaches. High in-line motion amplitudes, characterized by motion parameters, show highly nonlinear dynamics involved in drag force development and system stability. Results from previous studies show that neural network models closely replicate experimental drag forces but are affected by actuator noise and other sources of interference and error, while parametric modeling techniques may give a more concise representation that is also easier to interpret using physical principles. Phase plane portraits and Lyapunov exponent analysis confirm chaotic behavior under high in-line motion amplitudes that highlights the complexity of force interactions under such motion regimes. The analysis presented here highlights the importance of hybrid modeling approaches in predicting nonlinear hydrodynamic forces particularly for applications in ocean structural dynamics subjected to high-amplitude oscillations.
