In-house Research and Development
In-house Research and development
For several years now, we have internally set up, an R&D program in computational fluid dynamics. This program focuses on the development of numerical methods and solvers in the context of biphasic flow under OpenFOAM.
In order to equip ourselves with a powerful calculation tool for our activities in the maritime and offshore field, we have decided to develop marineFoam, an incompressible biphasic solver. marineFoam benefits from a VoF-type calculation method (different from that used by interFoam) and use the Ghost Fluid Method (GFM) in order to solve the problems of physical discontinuities when crossing an interface between two immiscible fluids (density and pressure).
For instance, in the maritime and offshore fields, this method drastically improves wave calculations by eliminating parasitic accelerations of the light phase. On the other side, the VoF formulation is based on the algebraic schemes (MCICSAM, CICSAM, HRIC, MHRIC, BICS …) for the transport of the phase indicator function. Note that we have coded our custom libraries from scratch. From the 6 DoF perspective, we also have developed many features for increasing robustness and capabilities of the standard 6 DoF solver of OpenFOAM. For example, we use the Adams Bashforth Moulton predictor-corrector method for calculating 6 DoF motions. We also work in the field of actuator disk theory for propeller simulations.
Finally, we also promote these developments for other sectors of activity such as nuclear.
The Ghost Fluid Method (GFM)
Fedkiw et al. (1999) initially proposed the GFM for sharp density handling in the context of compressible flows. The method has been subjected to various extensions (Huang and al. (2007), Vukčević et al. (2017)).
The GFM is a powerful way to solve the problem of spurious air velocities occurring due to the imbalance between pressure and density gradients during the pressure-velocity coupling. The method allows us to obtain a sharp dynamic pressure and density field across the free surface (as we can observe on this video) while the velocity field remains continuous. This done numerically by imposing jump conditions at the interface leading to interface corrected interpolation schemes.
The GFM also drastically improves wave propagation simulations.
6 DoF motion improvement
For ship design and complex marine simulations having a robust and powerful 6 DoF solver is very often required. Thusly, we have increased the capability of the six DoF rigid body solver of OpenFOAM. As mentioned above we use for instance a predictor-corrector method for integrating forces and moments. The video shows the simulation of a self-propelled ship in the head sea with 3 DoF motion. This special motion solver allows us to combine sliding mesh, deforming mesh, and moving mesh technique at once while handling hierarchies between two bodies.