Modelling flow-induced vibrations of gates in hydraulic structures

The dynamic behaviour of gates in hydraulic structures caused by passing flow poses a potential threat to flood protection. Complex interactions between the turbulent flow and the suspended gate body may induce undesired vibrations. This thesis contributes to a better understanding and prevention of gate vibrations by employing a variety of computational approaches. Simulations with the finite-element method are used to analyse the fluid-structure interaction of a new underflow gate type which exploits leakage flow to temper the excitation. A physical model experiment of the same configuration confirms this beneficial effect for a wider range of conditions. Furthermore, an outline of a control system is given that is based on data from acceleration sensors installed on the gates. It is shown how this system can be trained to classify future states and thus support operational decisions that avoid critical vibrations. Moreover, evolutionary computing is explored as a system identification tool for dynamical systems. The differential evolution algorithm was applied to recover the coefficients of several non-linear motion equations of self-excited oscillations from time signals. The collective results from these different methods help to eliminate problems related to flow-induced gate vibrations and so increase water safety.