####
SHEN, Y. T. & DIMOTAKIS, P. E. 1989 "Viscous and Nuclei Effects on
Hydrodynamic Loadings and Cavitation of a NACA 66 (MOD) Foil Section,"
J. *Fluids Eng. ***111**, 306-316.

*Abstract*

A series of experiments has been conducted on a two-dimensional NACA
66 (MOD) foil to examine the effects of viscosity and nuclei on cavitation
inception. In this paper the main discussions center on two foil angles
having different types of pressure loadings to represent a propeller blade
section operating at design and off-design conditions. At one degree design
angle of attack the foil experiences a rooftop-type gradually varying pressure
distribution. At three degrees off-design angle of attack the foil experiences
a sharp suction pressure peak near the leading edge. Cebeci's viscid/inviscid
interactive code is used to compute the viscous scale effects on the development
of the boundary layer, lift, drag and pressure distribu-tion on the foil.
Chahine's multibubble interaction code is used to compute the effect of
nuclei, test speeds, foil size and foil surface on traveling bubble cavitation.
Both computer codes are found to agree satisfactorily with the experimental
measurements reported here. Two assumptions commonly used to predict full
scale surface cavitation from model tests are examined experimentally and
theoretically. The first assumption states that cavitation inception occurs
when the static pressure reaches the vapor pressure. On the contrary, the
experiments showed that the water flowing over the foil surface sustained
significant amounts of tension during inception of midchord bubble cavitation
as well as leading edge sheet cavitation. The second assumption states
that there is no scale effect on the values of negative minimum pressure
coefficient. In the case of a rooftop-type pressure loading, the second
assumption is supported by the pressure numerical calculations. However,
in the case of a pressure loading with a strong suction peak near the leading
edge the value of negative minimum pressure coefficient is as much as 12
to 15 percent lower on a model than at full scale.