/ 270° \ 



Fig. 5- Polar torque diagram for an early 

 Savonius rotor. 



starting point . Experimental equipment has not 

 yet been obtained to quantitatively evaluate rotor 

 performance much below 0.05 knots. However, some 

 notion of the significance of rotational varia- 

 bility can be drawn from Fig. 6 which represents 

 considerable averaging from an early set of rotor 

 tests. These results indicate that for speeds 

 greater than 0.05 knots the rotational variation 

 is not large, perhaps less than 10$. This has 

 been generally borne out by tests with other 

 rotors although, as might be expected, the toler- 

 ances and condition of bearings becomes signifi- 

 cant at these low speeds. Obviously the effect 

 of rotational variability can be largely elimin- 

 ated if output is calculated in multiples of one 

 revolution of the rotor. 



REFERENCE ROTOR PERFORMANCE 



Rotor unit CS-2 (Fig. 2) was treated as a 

 reference for model comparison. CS-2 was towed 

 in tandem during practically all of the runs made 

 with the other models . Steady state analysis of 

 records was quite uniform, being based on the 

 time required for 5 to 10 revolutions taken at 

 least 15 revolutions after the carriage hcl been 



27 

 26 

 25 



COVERAGE ANGULAR SPEED DURING 15° OF ROTATION 

 C = MINIMUM AVERAGE ANGULAR SPEED 



DATA COURTESY 

 OF MAR)NE ADVISERS 



60 90 120 150 180 210 240 270 300 



ANGLE OF ROTATION FROM LOWEST SPEED POSITION (DEGREES) 



Fig. 6. Sample of angular speed variation during one revolution. 



360 



118 



