Issue 29

A. De Rosis et alii, Frattura ed Integrità Strutturale, 29 (2014) 343-350; DOI: 10.3221/IGF-ESIS.29.30 348 whereas the density is set to  s =8. No structural damping is prescribed. Beams are clamped at the walls of the channel. By varying the relaxation frequency, different Reynolds number Re are achieved, i.e. Re =10, 25, 50. In Fig. 6, the time history of the horizontal component of the tip displacement is depicted for the bottom beam. Notice that the no symmetry breaking is experienced in the simulations. Figure 6 : Two slender beams obstructing a flow: time history of the horizontal component of the tip displacement at different Reynolds number, Re =10 (red), 25 (green), 50 (blue). As it is possible to observe, as Re grows, the amplitude of the displacement tends to reduce. Such behavior has to be addressed to the fact that the viscous-dependent shear component of the stress induced by the flow upon the beam tends to vanish for progressively higher values of Re . On the other hand, a more turbulent flow induces more complex oscillation for the beam. In order to assess the effectiveness of the coupling algorithm to properly enforce the interface conditions, the authors define an interface energy rate at the fluid–structure interface that is computed from both the fluid, J f , and the structure, J s , sides as the product between forces and velocities computed from the fluid and solid sides, respectively. As shown in Fig. 7, the two rates lie on the diagonal of the graph, thus showing values relatively close each other. Only a part of the graph moves apart from the diagonal, which corresponds to the beginning of the simulation when the velocity profile suddenly impacts the beams. Finally, some snapshots of the velocity field are given in Fig. 8. Figure 7 : Two slender beams obstructing a flow: J f vs J s at different Reynolds number, Re =10 (red), 25 (green), 50 (blue). C ONCLUSIONS n this paper, a numerical framework able to simulate fluid-structure interaction problems has been presented. Specifically, the fluid domain is solved by the LB method, whereas structure dynamics is predicted via the non-linear TDG method. The IB method is adopted to handle fluid-structure interface conditions. The moving free surface is treated by a VOF-based method. All the ingredients described are coupled in a global algorithm that has been applied in the two different test cases presented. In the former, a dam break phenomenon is simulated by adopting a free surface model. Present findings are compared to experimental data, showing a very close agreement. In the latter, a constant uniform rightward velocity profile invests two slender beams and scenarios characterized by different values of the Reynolds number are dissected. The effectiveness and the accuracy of the proposed approach are assessed by computing the interface energy rates from the fluid and solid sides, which show very close values. Based on numerical results, the I