Free Surface Effects tn Hull Propeller Interaetton 
piece (see Fig. 2) was fitted to the propeller hub. The measured 
torque was corrected for bearing friction determined by replacing the 
propeller temporarily by a dummy hub. No ''dummy hub correction"! 
was applied to the measured thrust. 
The self-propulsion points were determined by the so-called 
British method, i.e. for each test run the towing speed and propeller 
rate of revolutions were preset while thrust, torque and residual tow- 
ing force were the quantities to be measured when the steady state 
condition had been reached, For each Froude number investigated, 
five to eight test runs at the same towing speed but varying rates of 
revolution were conducted to cover a wide range of propeller loading 
aroundthe ship self-propulsion point (and usually extending up to and 
beyond the model self-propulsion point). 
There was some indication (a characteristic knocking sound 
familiar from the previous open water tests) of mild ventilation at 
the highest propeller loadings encountered in the self-propulsion tests. 
However, there was no visible effect on the measured thrust and 
torque values. 
A.4. Wake measurement 
A set of standard Kempf & Remmers four-bladed wake wheels 
was used to measure the nominal wake in the propeller plane behind 
the hull in both forward and reverse motion, The diameters of the 
wheels available ranged from 40 to 220 mm in steps of 20 mm, The 
wheels were designed to yield directly the circumferential average of 
the axial flow velocity at the wheel radius. There was provision for 
turning the wheels around by 180 deg on their axis to ensure that the 
direction of flow relative to the blades was the same for both forward 
and reverse motions of the model (thus requiring only one set of cali- 
brations). 
The wheels were first calibrated in open water at a submer- 
gence of 150 mm (identical to that used for the model wake measure- 
ments) by means of a special towing device also supplied by the manu- 
facturer. In principle, the calibration curves (i.e. wheel rate of 
revolution as a function of towing speed) should have been linear. In 
practice, a few wheels showed pronounced nonlinearities and even 
mild discontinuities at some speeds, presumably due to flow instabi- 
lities. However, all calibrations were highly repeatable. 
For the actual wake measurements, the wheel towing device 
was mounted rigidly to the inside of the model with only its axis pro- 
jecting out of the stern tube on to which each respective wake wheel 
1885 
