2021 Volume 49 Issue 11 Pages 644-655
In this paper, we have conducted a wall-resolved large-eddy simulation(WRLES)for a cavitating flow around a Clark Y-11.7%hydrofoil by using two kinds of homogeneous cavitation model, one of which is a typical one-equation model by Okita and Kajishima(OK model)and the other is a four-equation model called multi-process(MP)model. First, we discussed the validity of the present WRLES for the non-cavitating flow, and a reasonable result was obtained under the given boundary condition. Next, it was confirmed that OK model based on the WRLES successfully reproduced an unsteady sheet/cloud cavitation with shedding of a sheet cavity while MP model did not simulate the sheet/cloud cavitation, i.e., the latter model just showed the unsteady change of the sheet cavity length without the shedding. The reason why such sheet/cloud cavitation is not reproduced in MP model might be caused by that the position of the sheet cavity front(inception point)is downstream compared with OK model, whose characteristic in MP model is closer to experiments. Namely, the sheet cavity front in the case of MP model appears downstream from a position of the suction peak pressure while the front position in the case of OK model corresponds to the suction peak as same as other homogeneous cavitation models. Taking account of that the inception point has been experimentally observed at a laminar separation or turbulent transition point appearing downstream from the suction peak, MP model may have a potential for better prediction of the inception point.