Credibility of wave breaking computations by Volume of Fluid RANS codes

Wave breaking is a frequent phenomenon in steady ship wave patterns. In ship wave computations using Volume of Fluid RANS codes common today, wave breaking seems to be predicted naturally and may look realistic. However, this paper shows that the computed flow has some quite unphysical features.
Computations have been made for the ‘Duncan foil’ case, a submerged foil at incidence that generates a wave pattern with a spilling breaker. Like in the experiments, the predicted flow shows a strong decrease of the wave amplitude aft of the breaker. However, this appears to be caused by a recirculating cloud of air with 15% of water, located above the wave surface and exerting a pressure on it due to its weight. There is a superficial similarity with earlier models of spilling breakers, but the air content is far too large. For plunging and spilling breakers in ship wave patterns, the same phenomenon is predicted, with the ‘breaking’ located entirely above the wave surface. A suspected cause of this deviating result is the absence of turbulent mixing of water and air in the formulation.
Until improvements have been found and further validation has been done, the physical realism of VoF RANS predictions of steady breaking waves seems quite limited.