The overview of the technology in the Naval Hydro Pack that we use for our simulations can be found in this document, while a summary of unique capabilities is given below.
We believe in integrated approach for CFD software development, where people developing the code actually test the code. There is no single best tool for every task (if you don’t trust us, try to open a bottle of wine with a hammer). This is why we have different interface capturing schemes suitable for different kinds of problems:
- Algebraic Volume-of-Fluid method
- Implicitly redistanced Level Set method
- Geometric Volume-of-Fluid method
The Ghost Fluid Method (GFM) enables infinitesimally sharp interface due to numerically consistent treatment of free surface jump conditions. That’s a complicated way of saying that we take into account the jump in density and pressure gradient at the free surface in a consistent manner, which solves the issue with spurious velocities in air near the interface.
People have been running numerical simulations in potential flow decades before the CFD simulations. We exploit this fact by coupling our fully nonlinear, two-phase and turbulent CFD simulations with potential flow solutions via:
- SWENSE solution decomposition where we solve only for the perturbation around the lower fidelity potential flow solution
- Domain decomposition based on relaxation zones to get rid of the wave reflection at the farfield boundaries
We also have the Higher Order Spectrum (HOS) method for extremely efficient nonlinear propagation of directional sea states. The method takes into account nonlinear wave modulation and wave-wave interaction in a potential flow framework, which is then coupled to viscous, two-phase CFD simulations.
Want to know how we managed to obtain three times the speed-up for seakeeping simulations without sacrificing the accuracy?
The theory is explained in the Enhanced hydro-mechanical coupling paper, but it’s actually simpler to explain it here briefly. After each pressure correction equation, simply solve the 6 DOF equations. Why? Because even without moving the grid, fluid pressure and velocity boundary conditions are tightly coupled. Again, something we learned from the potential flow world.