461 



xlO 

 6 



(a) BEFORE EXPERIMENT 



(b) AFTER Experiment 

 FIGURE 13. Impression of bent trailing edge. 



important for marine propellers where the scale 

 ratio between a full scale propeller and its model 

 is large. Sometimes this ratio exceeds 30. As 

 mentioned above , while the effect of the velocity 

 difference is very large, we can still make a model 

 test with the same tip speed as full scale by 

 increasing the revolution of the model propeller. 

 However it is very difficult to reduce the scale 

 ratio of chord length. 



Experimental verifications on this problem are 

 also very poor. Thiruvengadam (1971) made his 

 erosion tests using two chord lengths, 1.5 and 3 in. 

 His result shows that the erosion intensity increases 

 proportional to the chord length. The result 

 obtained in the present test is shown in Figure 15. 

 In the present tests the erosion intensity increases 

 proportional to the square of chord length. The 

 effects of hydrodynamic factors such as cavitation 



20 HO 70 100 

 Velocity (m/s) 



FIGURE 14. MDDR vs. velocity 

 (NACA 0015 : H2102-2, C = 30 mm, 

 a = 4 deg.) (NACA 16021 : H2102-2, 

 C = 40 mm, a = 4 deg.). 



1 



0,7 



O.iJ 

 0,3 



20 40 70 100 

 Chord Length (mm) 



FIGURE 15. MDDR vs. chord length 

 (NACA 0015, H2102-2, a = 4 deg., 

 V = 35 m/s) . 



number, velocity, and chord length can be explained 

 universally by a model of erosion mechanism. The 

 details of this model will be given in Section 5. 



Air Content 



The effect of air content was examined using the 

 NACA 16021 foil section results. The air content 

 was controlled as follows. As a pretreatment , the 

 water was degassed to about 8ppm in a vacuum chamber 

 and introduced into the cavitation tunnel. Then a 

 certain amount of air was injected into the tunnel 

 through an injection port before the test. In this 

 case the ratio of gaseous air to total air content 

 is much greater than found in ordinary water where 

 the amount of air is an order of parts per million 

 of total air content [Ahmed and Hammitt (1969) ] . 

 With increase of air content the value of MDDR 

 decreases as seen in Figure 15. This tendency 

 agrees with the test results of SSPA [Lindgren and 

 Bjarne (1974)] and those of Stinebring et al . 

 [stinebring et al. (1977)]. The reason is attributed 

 to the damping effect of air in a collapsing cavity 

 bubble, attenuation effect of tiny air bubbles to 

 shock wave, or a combination of both. 



Material Properties 



The effects of material properties on erosion are 

 usually tested by accelerating devices such as 

 vibrators, rotating discs, water jets etc. Summa- 

 rizing these results, Heymann has made the chart 

 shown in Figure 17 where the hardness of the 

 material was taken as a factor governing the erosion 

 [Heymann (1969)]. As seen in the figure the slope 

 differs according to the material group, namely 

 the slope of the steel group is steeper than that 

 of aluminum and copper and brass group. This implies 

 that the erosion resistance cannot be fully repre- 

 sented by hardness alone. Thus other material 

 properties such as strain energy absorbed to material 

 (engineering strain energy) [Thiruvengadam (1966)], 

 ultimate resilience [Hobbs (1966)], or their com- 



