Because of the speed squared term in the relation, any constant or variable 

 error in p, p , or V will produce a speed effect on the calculated value of C . For 



OK p 



example, if the static pressure, p as measured is greater than actually occurred, 

 then the measured value of p - p will be decreased over the range of speeds by a 

 given increment. When nondimensionalized by V , a monotonically decreasing pressure 

 coefficient will result over the given speed range. 



Errors in measurement of p were possible due to the many corrections performed 

 before the final value was reached. As stated previously, zeros were taken when the 

 propeller was stationary. The value was then corrected for wave height deviations 

 from zero and the hydrostatic correction to the propeller axis based on the angular 

 position of the propeller. Corrected static zeros were compared over twenty range- 

 of-speed runs with variations observed from 0.02 to 0.07 psi (0.14 to 0.48 kPa) . No 

 correlation of speed with pressure variation was observed. This variation was within 

 the expected gage error, and did not approach a p error of 0.2 - 10 psi (1-69 kPa) 

 which is necessary to produce the observed speed effect. The small variation in p 

 values demonstrated the proper zero correction, because each zero collection was 

 performed with the propeller at a different angular position. 



Errors in the measured pressure p were possible due to uncertainties in the 

 loading corrections. As stated earlier, some gages, primarily on Propeller 4718, 

 exhibited a sensitivity to blade loading. To correct the data for this effect, the 

 test conditions were repeated with tape over the gages (which then measured the 

 cavity air pressure) to desensitize the gage output to pressure. The measured load- 

 ing effect was subtracted from the measured pressures. The loading effect was found 



to be sensitive to advance coefficient, but relatively insensitive to Reynolds 



2 

 number or speed when nondimensionalized by l/2pV . Because the loading effect was 



independent of speed, it was thought not to be the cause of the observed speed 



effect. 



Correlations between gages with large loading and speed effects indicated a 



possible cause for some of the large speed dependency. A measure of the relative 



loading effect was represented by the quasi-steady loading correction Q, which is 



the slope of the C versus J curves at design J shown in Figure 7. Figure 15 shows 

 pL 



the magnitude of the loading correction Q against the speed dependency, as the 

 difference in pressure coefficient at two inflow speeds. For Propeller 4718 it 



21 



