290 



FIGURE 5. Experimental 

 arrangement. 



AFT VIEW 



PROFILE VIEW 



that it has twice the blade thickness (and a slight 

 difference in pitch to correct for the added thick- 

 ness) . The principal design characteristics of 

 the propellers are -listed in Table 3. The propellers 

 were designed by lifting-surface methods and both 

 open water performance [Denny (1968) ] and field 

 point pressure measurements [Denny (1967) ] have 

 been reported. It should be noted that the theoret- 

 ical predictions of field point pressures agree 

 very well with the experimental measurements (at 

 design advance coefficient) and the same propeller 

 theory is applied in the present surface force 

 calculations. 



The Force Dynamometer 



A dynamometer was developed to measure the horizon- 

 tal component of the unsteady forces produced on 

 the half body by the propeller. The half body is 

 cantilevered from the afterbody on five (5) flexures. 

 Forces are determined by measuring the strain in 

 one flexure, while the other four flexures absorb 

 the vertical force and moments as illustrated 

 schematically in Figure 6. The measurement flexure 

 transmits vertical forces and moments with miminal 



stress while resisting a large part of the horizontal 

 force (calculated to be over 90 percent) . 



Two competing requirements governed the flexure 

 design - the need to resolve small forces and the 

 desire to maintain the natural frequency of the 

 flexure-half body system far above the propeller 

 excitation frequency. Also the flexure was expected 

 to experience large (static) forces arising from 

 flow misalignment and hydrostatic loading. 



From the relationships for stress and stiffness 

 of a simple cantilevered beam, it is known that 

 for a given force , the flexure should have a low 

 stiffness in order to produce maximum strain. This 

 in turn would require a small body mass to keep the 

 natural frequency high. However, if the body is 

 too small, the resulting propeller force signal 

 becomes difficult to retrieve in the presence of 

 background noise . Although sophisticated techniques 

 were employed to reduce electrical noise and boost 

 signal power, it was not possible to completely 

 eliminate mechanical noise generated by the rumbling 

 carriage. VJith these compromises in mind, the 

 flexure was designed for a frequency ratio of 0.5, 

 producing minimally acceptable stress levels of 

 1000 psi (6.9 pPa) for the one pound (0.454 kg) 

 force in this experiment. 



TABLE 3. PROPELLER GEOMETRY 



