456 



2 4 



Time (hr) 



FIGURE 4. Extended duration cavitation test [NACA 

 16021, Pure Al (H2102-2) , C = 40 mm, a = 4 deg.]. 



posed by one of the authors, seems to be more 

 suitable than MDP. The advantages of using MDD 

 are that it increases almost linearly over a wide 

 range of exposure time as well as the fact that MDD 

 corresponds to SR in the incubation period and to 

 MDP after long exposure. 



In the present tests , SR was measured to shorten 

 the testing time. Usually the test was completed 

 within 1 hour so the SR value coincides with MDD. 

 The degree of erosion after a long exposure can be 

 estimated using the measured SR. 



3. 



HYDRODYNAMIC CHARACTERISTICS OF CAVITATION ON 

 NACA 0015 FOIL SECTION 



measured. At the test condition 50 photographs 

 were taken to measure the cavity length. 



The results are shown in Figure 5 . As seen in 

 the figure, above a > 0.8 the distribution of cavity 

 length is characterized by a peak, but below a < 0.8 

 the fluctuation becomes so large that there is no 

 characteristic peak. For the supercavitation 

 condition (a = 0.45) the fluctuation is reduced and 

 a characteristic peak can again be observed. The 

 mean value of cavity length and its standard devia- 

 tion are shown Figures 6 and 7. The cavity length 

 increases linearly with smaller cavitation number, 

 and the standard deviation begins to increase 

 rapidly about a = 0.85 as clearly seen in the figure. 



It is well known that the cavity length of a 

 partially cavitated foil can not be determined 

 theoretically by linear cavity models. The cavity 

 length predicted by a closed type cavity model is 

 usually longer than the observed length. If we 

 adopt a open type cavity model. , the situation 

 becomes reversed and the predicted cavity length 

 becomes shorter than the observed length. Conse- 

 quently a half-closed type model is usually adopted, 

 but this model requires the opening of the cavity 

 end to be determined experimentally. 



In this study the cavity length was calculated 

 using the half-closed type model by Nishiyama and 

 Ito (1977) . This method is based on linear theory 

 using singularities (source and vortex) distributed 

 on the cavitated foil. The calculated results are 

 shown in Figure 7 where the opening 6e was system- 

 atically changed. The contour of 6e = coincides 

 with the closed cavity model. The circles in this 

 figure represent the "mean" value of the observed 

 cavity length. Using this mean value, the opening 

 Se can be calculated showing that 6e increases 

 with smaller values of a (see Figure 8) . 



Cavity Length 



Because erosion occurs at the collapsing point of 

 the cavities namely the end of the cavity, it is 

 important to know the cavity length for predicting 

 cavitation erosion. Therefore, prior to the erosion 

 tests, the cavity length and pressure distribution 

 along the back surface of the NACA 0015 foil were 



Pressure Distribution and Cavity Shape 



The theoretical pressure distribution and cavity 

 shape for the back side of NACA 0015 foil section 

 are shown in Figure 9 along with the corresponding 

 experimental result. Here the Nishiyama-Ito' s half- 

 closed model was used with the 6e values taken 



100 



50 



»■= 1.051 

 Mean 



JtliL 



50 100 150 

 Cavity Length (7.Chord) 



100 



:: 50 



r 



a= 0.500 



Mean 

 — I- 



M-TL. iii>JjTri-n 



FIGURE 5. Fluctuation of cavity length (NACA 

 0015, o = 4 deg., V = 35.9 m/s). 



50 100 150 



Cavity Length (%Chord) 



100 



50 



r 



o-= 0.830 



Mean 



oM. 



EbzL. 



50 100 150 



Cavity Length (%Chord) 



100 



50 



a-=0.4'j6 



Mean 



rh„n. 



-Mh. 



50 100 150 

 Cavity Length (ZChord) 



