Titolo della tesi: Development of a density-based solver for compressible turbulent flows in complex geometries
This work presents the development, implementation, and validation of a novel
density-based solver for compressible turbulent flows in complex geometries. The
research focuses on creating a computational fluid dynamics tool capable of accurately
simulating aerospace applications with diverse flow regimes. The solver
employs energy-preserving numerical schemes on unstructured grids, extending
state-of-the-art methodologies to practical engineering problems. A hybrid approach
combining these schemes with shock-capturing techniques enables accurate resolution
of both smooth and discontinuous flow features. The implementation includes both
Reynolds-Averaged Navier-Stokes and Large Eddy Simulation turbulence modeling
capabilities, validated through canonical test cases and complex configurations. The
solver demonstrates excellent parallel performance on high-performance computing
architectures, with efficient CPU and GPU utilization. Comprehensive validation
against analytical solutions, experimental data, and numerical benchmarks confirms
the solver’s accuracy across subsonic, transonic, and supersonic flow regimes. The
resulting computational framework bridges the gap between academic research and
industrial applications, providing a valuable tool for both fundamental fluid dynamics
research and practical aerospace engineering problems.