390 



1.5 



1.0 



0.5 



"^^ 



* t 





(b) a =0.052 rod 



0.2 0.4 06 



3.0 



2.5 



0.2 0.4 



0,6 



0.8 1.0 



y/h 



FIGURE 8. Spanwise variation of incipient cavitation 

 numbers for the 0„ profile. 



it reaches 0.105 rad, but become smaller at a = 

 0.157 rad. 



At a's not smaller than 0.105 rad, fixed cavita- 

 tions occur in the boundary layers at positions 

 very close to both side walls at kd's much greater 

 than local |cpmin| 's. At the same time a zone of 

 cavitation widens spanwise beyond each boundary 

 layer with the inception so that detection of 

 inception becomes difficult in the region neighbor- 

 ing both boundary layers on the side walls. This 

 is the reason the lack of points between y/h = 0.025 

 - 0.3 and 0.7 - 0.975. Frequency distributions of 

 cavitation occurrences analyzed by using high speed 

 motion pictures for 1 second illustrate those facts, 

 as can be seen in Figure 9. 



In free streams with shears made by the grids 

 No. 1 and No. 2, kdi's almost equal or are a little 

 larger than local | Cpmin | ' s . They vary spanwise 

 under the influences of the flow shears in the 

 core and the boundary layers on both side walls , 

 and the accompanied secondary flows , except at 

 a = 0.105 rad, which indicates that these free 

 streams are rich in cavitation nuclei. At a = 0.105 

 rad, kdi is a little smaller than | Cpmin |, which 

 can be assiomed to be due to cavitations changing 

 from traveling to fixed, as mentioned in the next 

 section. 



Differences between kdi's and |cpmin|'s in the 

 boundary layers are larger than those in the case 

 of no grid on the low-speed side, but are the 

 contrary on the high-speed side, due to the 

 secondary flows induced by the flow shears in the 

 cores. The above-mentioned effect is most remark- 

 able at a = 0.105 rad: kdi's in the boundary layer 

 on the low-speed side in cases of the shear grids 

 are larger than those not only in the case of no 

 grid but also |cpmin|'s in the boundary layer, 

 though only by a little. The mechanism causing 

 the effect has been examined by measuring spanwise 

 variations of static pressures on three points near 

 the leading edge in the boundary layer on the low 



speed side at the attack angle of . 105 rad in the 

 case of the grid No. 2. It was confirmed that the 

 detected incipient cavitation number, 2.53, in the 

 boundary layer lies near the largest absolute value 

 of the pressure coefficient based on the local 

 velocities in the zone between 3 and 5 mm from the 

 side wall. However, measured velocities in the 

 zone are not very reliable. Symbols A in Figure 

 7 show kdi's when the hydrofoil has a tip clearance 

 of about 0.1mm on the high-speed side in the case 

 of grid No. 1. It was found that effects of a 

 boundary layer are weakened by tip clearances, 

 especially at large angles of attack, although 

 another cavitation occurs at the tip clearance. 



08 Profile 



At angle of attack, traveling cavitations 

 occurred and kdi almost coincide with | Cpmin | in 

 the case of no grid, but were larger than the latter 

 in the case of grid No. 1. The difference decreases 

 spanwise toward the high-speed side, in correspon- 

 dence with the size distribution of cavitation 

 nuclei. At angles of attack larger than rad, 

 however, fixed cavitations occurred and kdi's were 

 much larger than measured | Cpmin | 's because of the 

 lack of a piezometer hole at the position of the 

 largest | Cpmin | , which is closer to the leading 

 edge than the closest hole at 3% chord. At a = 

 0.052 rad, in the case of the grid No. 1, another 

 cavitation of the traveling type appears around 

 the position of the measured second | Cpmin | and the 

 kd almost coincided with the measured | Cpmin]. kdi's 

 in the case of grid No. 1 were smaller than those 

 in the other case on the high-speed side. The 

 discrepancey can be surmised as due to the discrep- 

 ancy between structures of laminar separation 

 bubbles just behind the leading edge in the two 

 cases because of the difference between turbulence 

 levels. At a = 0.105 rad, kdi in the case of the 

 grid No. 1 was larger than in the case of no grid 

 in the core of free stream, but was the opposite 

 in the boundary layer on the high-speed side wall. 

 In the case of no grid, cavitations with long and 

 wide zones occurred in the boundary layers on both 

 sides close to the side walls and the leading edge 

 as with the Clark Y 11.7 profile. 



Location of Incipient and Developed Cavitations 



Spanwise variations of positions of cavitation 

 inception and front and rear edges of (time) average 

 zones of developed cavitation are shown in Figures 

 10 and 11 for the profiles Clark Y 11.7 and 08 

 respectively. Also are shown spanwise variations 



0.4 



0.3 



"^ 0.2 



0.1 



FIGURE 9. Frequency q 

 distribution of cavi- 

 tation occurrence. 



0.2 04 0.6 0.8 1.0 

 y/h 



