appeared that many of the gages with large loading corrections also had large speed 

 effects. Also, most of these gages were the back-mounted type used in the thicker 

 sections of the blade. Propeller 4679 did not have as large a speed effect or as 

 great a loading effect as Propeller 4718, but showed similar correlation between the 

 two effects, and indicated greater problems with the back-mounted gages. 



From this observation it was thought that some errors existed in the loading 

 correction procedure. The most probable error was a coupling effect between the 

 gage calibrations for loading and pressure. When the loading sensitivities were 

 measured, the gage holes were taped over, producing approximately constant atmos- 

 pheric pressure acting on the gage. During pressure measurements, additional non- 

 atmospheric pressures acted on the gage, therefore, the same conditions occurring 

 during the loading effect measurements were not reproduced. If the loading effects 

 were caused by gage diaphram deflection, produced by displacements of the blade in 

 contact with a gage, then that effect could be dependent upon the initial deflection 

 of the gage, before the blade was strained by the loading. The dependence of the 

 loading sensitivity on gage deflection or pressure was checked in the laboratory by 

 hanging a weight from the end of the blade while applying known pressures to the 

 gage of interest. The gage output due to loading was independent of the pressure on 

 the gage. The lack of interaction effects between loading and pressure eliminated 

 the loading effect as a possible cause of the observed speed effect. The correlation 

 between loading effects and speed effects, if not due to instrumentation, could be 

 caused indirectly by systematic real flow effects for the speed-aff ected gages and 

 may be influenced by their location in the thicker mid-chord region of the blade. 



In future pressure measurement tests, closer attention should be paid to the 

 loading effects on the gages. It appears that on both propellers, the coverplate 

 configuration produced little or no loading effect, while the back-mounted configura- 

 tion produced large loading effects on Propeller 4718, and lesser effects on 

 Propeller 4679. As stated earlier, it is suspected that the back-mounted configura- 

 tion is very sensitive to slight variations in installation. In the future, a new 

 gage mounting configuration could be designed incorporating the advantages of both 

 types. 



22 



