604 



HYDRODYNAMICS IN SHIP DESIGN 



Sec. 70.16 



(e) Straight in the disc plane, not coinciding with 

 a radius, from the hub radius to the tip radius 



(f) Curved, usually with a sweep-back toward 

 the tip that is opposite to the direction of rotation. 



Whatever the shape of the locus on the propeller 

 which is not to pass simultaneously across the 

 basic reference trace abaft the hull ending, the 

 former is drawn on a sheet of thin transparent 

 material to the same scale as the reference trace. 

 In Fig. 70. C the transparent material is shown 

 as rectangular in outline and the selected-skew- 

 back locus is drawn in its proper position with 

 reference to a straight radial line tangent to the 

 locus at the hub radius. To the transparent sheet 

 there are added concentric circles at 0.2/?, 0.3i2, 

 and so on. A pin is inserted through both the 

 transparent overlay and the under sheet upon 

 which the skeg ending is drawn, at the propeller- 

 shaft axis, so that the overlay may be rotated to 

 simulate the rotation of the propeller. An arc 

 is added on the under sheet, just outside the tip 

 circle in the figure, to indicate the angular amount 

 by which the overlay is rotated from an assumed 

 zero position. 



The overlay is turned in the direction of pro- 

 peller rotation until the locus on it crosses the 

 basic reference trace at some selected fraction of 

 the maximum radius, say at 0.2/2. The angular 

 position of the overlay is then noted, following 

 which it is turned until the locus crosses the 

 reference trace at the next selected radius fraction, 

 say 0.3/2. The angular position is again noted. 

 By taking the angles between successive angular 

 positions, set down as first differences in the table 

 on Fig. 70.C, it is possible to determine at a 

 glance whether the successive differences between 

 a group of adjacent angles are small or appreciable, 

 or whether they are equal or unequal. A group 

 of equal and appreciable differences indicates 

 successive crossings of the locus and the reference 

 trace at equidifferent angular and time intervals. 

 A group of zero differences indicates a simul- 

 taneous crossing over the corresponding portion 

 of the blade radius. This simultaneous crossing 

 is almost invariably to be avoided. 



A limited study of propellers with swept-back 

 blades that have performed well in service 

 indicates that successive crossings of the leading 

 edge are spaced at more nearly uniform angular 

 intervals than the crossings of any other known 

 locus. Therefore, until a better locus is found it 

 appears acceptable to take the leading edge of the 



-projected outline as the one whose radial segments 

 should swing successively and uniformly into the 

 high-wake region. 



It is not feasible, however, by any method as 

 yet known, to fashion a suitable projected-blade 

 outline, with fair root-to-tip characteristics, by 

 working from the leading edge. Experience indi- 

 cates that the locus of the midlengths of the 

 expanded blade sections is the most suitable 

 construction line in which to introduce the 

 skew-back. This midlength locus, to which are 

 applied the half-chord expanded lengths of the 

 forebodies of the blade sections at the various 

 radii, is then adjusted in shape until the leading 

 edge resulting from this construction gives the 

 desired equidifferent angular intersections. This 

 is determined by the method diagrammed in 

 Fig. 70. C, substituting the leading edge for the 

 purely schematic rotating locus shown there. 

 At or near the tip it is expected that the angular 

 differences will increase because of the large- 

 radius curves used to form the tip profile. The 

 midlength locus is preferred over the locus of the 

 chordwise positions of the maximum thickness 

 of each section because the former is usually 

 made tangent to the pitch reference line at the 

 hub and it has less curvature at the outer radii. 



Until more is known concerning the controlling 

 locus, it appears wise, if possible, to give definite 

 angular separation at the successive radii to both 

 the midlength-of-chord locus and to the position- 

 of-maximum-section locus. In addition, both 

 should be fair from root to tip of the blade, as 

 should the leading and trailing edges. 



Good values for the skew-back at the tip, 

 measured as indicated in Fig. 70. of Sec. 70.36 

 and by Fig. 78. L, lie between 20 to 25 per cent 

 of the maximum chord length Cmsi of the blade. 



The designer is cautioned, on a moderately or 

 heavily loaded propeller at least, never to apply 

 an appreciable amount of skew to a screw- 

 propeller blade in a forward or ahead direction, 

 for the reasons given in Sec. 70.44. 



For a given maximum thickness tx at each 

 section, a blade with large sweep-back has a 

 smaller thickness ratio tx/c because the lengths 

 of the sections are usually greater in the circum- 

 ferential direction of flow across the blade than 

 they would be in a blade with no sweep-back. 

 This thinness is generally a help in deferring 

 cavitation. 



Theoretically the dynamic ram pressure built 

 up on the extreme leading edge of a fast-running 



