We investigate two strategies to reduce the computational cost of forced-motion simulations in obtaining flutter derivatives. First, the multi-frequency approach, consisting of exciting a cross-section at more than one frequency at a time, is analysed in detail. Several LES-type investigations are carried out to determine the limits of this methodology, comparing the results with the classical single-frequency simulations. Second, we explore the possibilities of a URANS-based approach. Both turbulence models are developments in the context of the finite element method. Our strategies are applied to find the flutter derivatives of two cross-sections: a rectangular cylinder with proportions 5:1 and a generic streamlined bridge cross-section. Here we only include an excerpt of the results for the latter shape. The results obtained from the multi-frequency simulations show excellent agreement with the ones obtained through classical means. However, when it comes to decreasing the computational cost, the URANS simulations are considerably more efficient, remaining a viable option for estimating the flutter derivatives. We highlight the possibility of combining these strategies, potentially leading to a computational cost of less than 1% of the original effort.
«
We investigate two strategies to reduce the computational cost of forced-motion simulations in obtaining flutter derivatives. First, the multi-frequency approach, consisting of exciting a cross-section at more than one frequency at a time, is analysed in detail. Several LES-type investigations are carried out to determine the limits of this methodology, comparing the results with the classical single-frequency simulations. Second, we explore the possibilities of a URANS-based approach. Both turb...
»