oscillator which, in turn, was rigidly mounted on the towing carriage. 
This arrangement enabled the model hull and the drive system to be dy- 
namically pitched together while maintaining independent support from one 
another. 
The sensing elements were flexures to which were bonded high- 
sensitivity, semiconductor strain-gage bridges. The design of these 
flexures has been described by Dobayeae There were three flexures, each 
of which measured two components of blade loading. Flexure 1 measured 
components Ee and Hho Flexure 2 measured components Be and M> and Flexure 
3 measured components Ee and M, (Figures 1, 5, and 6). An arrangement of 
three separate flexures rather than one to measure all components of blade 
loading was adopted Because it appeared to result in higher natural fre- 
quencies (Criterion 1), higher sensitivities (Criterion 3), and lower 
interactions (Criterion 4), than would have resulted had a single flexure 
been used. 
The flexures were mounted inside a propeller hub which was specifi- 
cally designed for these experiments (Figure 6). Only one flexure could 
be mounted at a time, because of space limitations, and this necessitated 
duplicate runs, as discussed later in the section on experimental condi- 
tions and procedures. 
The strain-gage bridges were excited by a common d.c. voltage 
source, transmitted through the sliprings on the propeller shaft. The 
constant-current excitation used by DObayae was not employed in the pre- 
sent experiment because it appeared to be too sensitive to temperature. 
The voltage output from the flexures (due to blade loading) was 
transmitted through the sliprings to individual amplifiers (NEFF 119-121). 
These amplifiers utilized field effect transistors to produce an extremely 
high input-impedance (100 megohms, minimum). This high impedance essen- 
tially eliminated slipring noise to the amplifier. The voltage signals 
were transferred across the sliprings in the presence of only a small 
amount of noise-producing current. The amplifiers used here had zero- 
phase shift qualities in the d.c. to 20 kilohertz range. They were chopper- 
stabilized to enable both the steady and unsteady signals to be recorded 
simultaneously. This signal-conditioning system was essentially the same 
as that used by Dope. 
iil 
