Unsteady Cavitation on an 

 Oscillating Hydrofoil 



Young T. Shen and Frank B. Peterson 



David W. Taylor Naval Ship Research and Development 



Center, Bethesda, Maryland 



ABSTRACT 



Bent trailing edges and erosion are often observed 

 on marine propellers and are attributed mainly to 

 unsteady cavitation caused by the non uniformity 

 of the flow field behind a ship's hull. In order 

 to improve the physical understanding of the 

 cavitation inception and the formation, of cloud 

 cavitation on marine propellers, a large two 

 dimensional hydrofoil was tested in the DTNSRDC 

 36-inch water tunnel under pitching motion. Fully 

 wetted, time dependent, experimental pressure 

 distributions were compared with Giesing's unsteady 

 wing theory. The influence of reduced frequency 

 and pressure distribution on inception was determined. 

 A simplified mathematical model to predict unsteady 

 cavitation inception, was formulated. Good corre- 

 lation between theoretical prediction and experi- 

 mental measurements on cavitation inception was 

 observed. The reduced frequency, maximum cavity 

 length, foil surface pressure variation, and time 

 sequential photographs were correlated with the 

 formation of cloud cavitation. A physical model 

 based on the instability of a free shear layer 

 defining a near-wake region provides a reasonable 

 explanation of the observed results. 



1 . INTRODUCTION 



Hydrofoil craft are typically designed to operate 

 both in calm water and waves; and marine propellers 

 normally operate in the nonuniform flow field 

 behind a ship. Unfortunately, due to the complexity 

 of the experiments, only a few experiments have 

 been specifically concerned with unsteady leading 

 edge sheet cavitation on hydrofoils and propellers, 

 Morgan and Peterson (1977) . It is the intent of 

 this paper to report the results of experiments 

 concerned with leading edge sheet cavitation on an 

 oscillating two dimensional hydrofoil. Following 

 a brief review of the most pertinent experimental 



data available in the literature, an analytical 

 method for the prediction of inception will be 

 developed and compared with the experimental data. 

 Once the cavity is present on the foil, cavity 

 instabilities develop due to the foil oscillation 

 and also due to the inherent instability of the 

 cavitation process. This general process of 

 instability in the leading edge sheet cavity is the 

 subject of this paper. 



It has been observed by innumerable investigators 

 that a leading edge sheet cavity can, under certain 

 circumstances, be quasi steady with relatively few 

 collapsing vapor bubbles to produce erosion. How- 

 ever, if a propeller blade enters a wake field, the 

 inception angle of attack at the leading edge may 

 not agree with the uniform flow inception angle. 

 In addition, the developed cavity may exhibit 

 instabilities not produced in uniform flow fields. 

 One form of cavity instability is manifest by the 

 shedding of a significant portion of the sheet 

 cavity. This shed portion appears to be composed 

 of microscopic bubbles and is commonly referred to 

 as "cloud" cavitation, van Manen (1962) . Cloud 

 cavitation is now considered to be one of the main 

 causes of erosion and bent trailing edges , Tanibayashi 

 (1973) . 



Model experiments have been performed by many 

 organizations in an attempt to simulate full-scale 

 wake fields in which propellers operate. One of 

 the first detailed experiments concerned with 

 unsteady cavitation was reported by Ito (1962). 

 These experiments were with pitching three dimen- 

 sional hydrofoils and propellers in a wake field. 

 A principle result directly applicable to the work 

 to be reported here was that the reduced frequency 

 had an important influence on the cavitation. He 

 also concluded that the "critical" reduced frequency 

 at which a leading edge sheet cavity broke up into 

 cloud cavitation was 0.3 to 0.4. His latter con- 

 clusion will be considered in more detail in the 

 context of the results to be reported here. 



A recent discussion of this subject was given 



362 



