A perturbation method and model order reduction for stochastic hydroelastic problems

This study aims to optimize computing time during numerical modal analysis of hydroelastic problems with uncertain parameters. We proposed a strategy that combines two methods adapted to hydroelastic problems. The first is the modal superposition method for reducing the finite element model order. The second one is a perturbation method that considers the design parameters uncertainties and evaluates without iterations the natural frequencies moments of the wetted structure. The proposed strategy was validated by applying it to the stochastic characterization of the vibratory behavior of a water storage tank. Results of the proposed strategy were compared with those from the classical one, which consists of modal analysis without model reduction combined with Monte Carlo simulation. The confrontation showed a good correlation between the results of the two strategies. However, as expected, the proposed one was considerably computationally time inexpensive.