of ~. Several passes down the towing tank were required in order to obtain 
a sufficient number of samples. After all the data had been sorted based 
on ~, sign of }, and tolerance, the cycles for each combination of jp, and 
sign of ~ were analyzed in exactly the same manner as the data for the 
steady-ahead condition at fixed jp. 
For unsteady-acceleration runs, the model speed V varied with time 
t. During an acceleration run, data, including a measure of V, were sam- 
pled and stored in the same manner as for the steady-ahead runs. 
Five values of V were specified for analysis. For each acceleration 
run and for each specified V, the computer selected the propeller revolu- 
tion which had the average value of measured V nearest to the specified V. 
However, because only one revolution at each specified velocity was ob- 
tained for a single acceleration run, each such run was repeated five 
times. This yielded five revolutions at each specified velocity. All the 
cycles for each specified V were then analyzed in exactly the same manner 
as the data for the steady-ahead conditions. 
Thus, the on-line analysis system yielded average wave forms and 
harmonic analyses of the average wave forms for steady-ahead conditions, 
for specified conditions of ~, sign of W during the dynamic pitch cycle, 
and for specified velocities V during the acceleration operation. However, 
these on-line results are preliminary because: 
1. They do not consider the interactions between the various load 
components. These interactions were determined during the static cali- 
bration of the flexures. 
2. They include the complete measured signals with no filtering. 
As discussed in the section on experimental results, some extraneous sig- 
nals near the natural frequency of the flexure being used appeared to be 
superimposed on the signals generated by blade loading. 
3. They include the effect of centrifugal and gravitational load- 
ing on the aluminum model propeller. The centrifugal and gravitational 
components of loading were measured separately during air-spin experiments, 
as discussed earlier. 
4. They do not have any corrections for the influence of the dyna- 
mometer boat. These corrections are discussed later. 
19 
