The minicomputer automatically performed the entire calibration in a consistent 



fashion, based on programed variations such as the number and order of pressure 



steps, and the averaging and settling time for each pressure reading. After 



calibration data were obtained, the computer calculated a straight line least-squares 



curve fit based on the recorded pressures for each gage, providing gage sensitivity 



and the standard deviation from the straight-line calculated values. This system 



permitted consistently run, quick calibrations conducive to statistical analysis and 



« 

 the identification of possible systematic errors. 



An exhaustive series of calibrations was conducted on Propeller 4679 to inves- 

 tigate possible systematic errors in the pressure measurement instrumentation. 

 Initial calibrations were conducted in the laboratory without the 1000 hp dynamometer 

 cabling and sliprings connected through the measurement system. The propeller 

 pressure gages were calibrated under conditions with both water and air in the 

 pressure tank, and in the gage cavities. The procedure for filling the cavities with 

 water involved injecting water mixed with a wetting solution through the gage hole. 

 The wetting solution eliminated the adhesion of air bubbles to the cavity interior. 

 The procedure was used throughout the experimental program to remove air from the 

 cavity. Combinations of air and water in the cavities and pressure tank had no 

 effect on the gage sensitivities. 



Calibrations were also conducted with the propeller and pressure tank mounted 

 on the 1000 hp dynamometer. This arrangement most closely resembled actual test 

 conditions by including the dynamometer cabling and sliprings in the calibrations. 

 The pressure tank was also designed to be rotated with the propeller on the dyna- 

 mometer shaft, allowing calibration to include the effects of centrifugal loading, 

 propeller drive motor noise, and slipring noise. Calibrations were conducted with 

 air in the pressure tank and gage cavities while the propeller was rotating at 300- 

 500 rpm, representing typical test rotational speeds. Rotation had no effect on the 

 gage sensitivities. Some additional noise, developed on selected gage signals, 

 attributed to drive motor transmission noise because of its dependency upon the FM 

 multiplexing frequency of the gage channels. Because the source of noise did not 

 influence the sensitivity of the gage, it was assumed to average out in the data 

 collection process. 



