•"'-' Sevik 



During the tests turbulence was generated by grids mounted 20 mesh sizes 

 upstream of the propeller. Two grids were used with mesh sizes of 4 inches and 

 6 inches respectively. The first had a solidity of 0.34 and was fabricated with 

 3/4- inch-diameter rods. The second had a solidity of 0.27 and consisted of 

 7/8-inch-diameter rods. Figure 9 shows the propeller photographed from up- 

 stream of the 4- inch grid in the water tunnel. 



[■■■■■■■■■■■■■■■■■■■■Ik 

 [■■■■■■■■■■■■■■■■■■■■■b 



[■■■■■■■■■■^^■■■■■■■■■■■i 



^■■■■■■■■■•li&flaBBaaHaHii 3 



Haaaaaaayjhiaaaaaaaaaaa 7 



maaaaaaaaaaaaaaBEsaKirv 



IJEDBaaBBBBBBBB«EBaai9, 



■■EaBaflaBBBBBBBBliB», 



VpBifiEilBBBBBBBBKBBW;^ 



k^^lHHGfllHBHHHHBflB^ 



Fig. 9 - Propeller as seen through a 4-inch 

 grid upstream of the propeller 



The test were performed at a velocity of 15 ft/ sec and at two advance 

 ratios. Initially, runs were made without grids in order to establish the noise 

 level of the balance and of the electronic data acquisition system. The grids 

 were then mounted in the tunnel, and the measurements were repeated. During 

 the entire test program the tunnel pressure was maintained at 20 psia, which 

 ensured absence of cavitation. At the completion of the program the balance 

 was recalibrated, and it was established that its characteristics were unaltered. 



The circuitry used for data acquisition is shown in Fig. 10. Three SKL 

 Model 308 A variable electronic filters were set up as a bandpass filter. Two, 

 used as high-pass filters, were cascaded to produce a 48-dB-per-octave rejec- 

 tion rate below the cutoff frequency. These filters eliminated the strong bal- 

 ance output at a frequency corresponding to the shaft rps. The cutoff frequency 

 of the high-pass filters were varied with the rps of the propeller shaft. For the 



306 



