Verification and Validation of CFD for Surface Combatant 5415 for Straight Ahead and 20 Degree Static Drift Conditions

摘要

Collaboration is described on verification and validation of CFD for surface combatant model 5415 for static drift β=0 and 20 degrees using recent experimental data for: forces and moment, wave elevations, and tomographic PIV planar and volume measurements of velocity, vorticity and turbulent kinetic energy (TKE), including analysis of vortex onset of separation, progression, instability and TKE budget. Results were obtained from five different solvers, which used different numerical and discretization schemes, free-surface and turbulence models and adapted grids ranging from 2.5M to 102M for β=0° and 4.6M to 250M for β=20°. For β=0°, resistance and wave-elevation predictions compare within 2% and 3.5% of the data, respectively. Solvers agreed with the data for the onset of the primary vortices, but showed large variation in their progression and decay. URANS turbulence models predicted premature decay of vortices, whereas DES predicted too strong vortical structures and low resolved turbulence. The primary vortices exhibited open-type separation. For β=20° forces and moment compared within 3.7% of the data, and wave-elevation within 8.5% of data. Simulations agreed with the data for the onset of primary vortices, but showed large variations for their progression and decay, and for the leeward sonar dome separation bubble and breaking waves. DES performed better than URANS for the prediction of vortex strength and TKE. The vortices show many open-, closed- or open-closed type separations. The primary vortices show helical mode instability, and the instability frequency for the sonar dome tip vortex at x/Lpp =0.4 compared within 11.3%D of the data. TKE budget revealed that the production occurs at the vortex inception and is transported by pressure or turbulent fluctuations. The finite-difference solver provided better vortex decay predictions than finite-volume solvers for β = 0°. Level-set and VoF provided similar wave elevation predictions except for breaking waves. The study indicates the need for more accurate turbulence closures. Future research should focus on investigation of: improved RANS models such as Reynolds stress transport model; and improved hybrid RANS/LES models to address the turbulence trigger and modeled stress depletion issues of DES.