The proposed session brings together the latest research and advances in fluid-structure interaction (FSI), uniting complementary perspectives from both detailed local-scale modeling and large-scale engineering applications. The aim is to gather a diverse community to exchange knowledge on a broad spectrum of approaches — from refined methods that explicitly represent fluid-structure interfaces (e.g., Arbitrary Lagrangian-Eulerian formulations) to more robust strategies such as immersed boundaries or porous media models, as well as hybrid couplings involving particles, finite elements, finite volumes, or Lattice Boltzmann methods. The scope encompasses a wide range of physical regimes — incompressible turbulent flows, multiphase flows, and compressible flows — interacting with various structural models, including cases with strong nonlinearities such as contact, impacts, and fluid-elastic instabilities. The session will address applications from multiple industrial fields, including power plants, biological systems, and offshore engineering. A distinctive focus will be placed on multi-scale modeling and upscaling strategies to deliver certified computational mechanics tools with quantified confidence levels for industrial end-users. This includes rigorous uncertainty quantification and propagation from the local to the industrial scale, as well as validation at every relevant scale through combined experimental and numerical studies. Key topics of interest include: - Efficient and robust partitioned coupling methods, extended to extreme-scale computing - Local-scale advanced modeling and its connection to homogenized or porous representations at large scale - Strategies for managing and validating nonlinear dynamic responses of immersed structures - Effects of turbulence, compressibility, wave propagation, and added mass/damping coefficients - Cross-community knowledge sharing on state-of-the-art solutions for complex FSI problems By bridging fine-scale analysis with industrial-scale applications, the session seeks to provide a platform where methodological diversity meets engineering reliability, fostering collaboration across theoretical, computational, and experimental domains.