Sec. 66.5 



STEPS IN PRELIMINARY DESIGN 



465 



(a) The proper relationship of speed and length 

 so that a hump in the hull-resistance curve is 

 avoided 



(b) A suitable balance between the added friction 

 drag on a too-long hull, with its extra wetted 

 area, and the added pressure drag on a too-short 

 one, with its sharper longitudinal curvature. 



G. S. Baker gives a dimensional formula for 

 determining the length Lpp between perpen- 

 diculars, namely 



Lipp — 2i'± 



V 



2+ V. 



A' 



(66.i) 



where Lpp is in ft, V is "the speed (in kt) for 

 average fine weather at sea," and A is in long 

 tons [NECI, 1942-1943, Vol. 59, p. 29]. Taking 

 first the sustained speed V of 18.7 kt for the ABC 

 ship, and the estimated displacement A of 

 17,300 t, Eq. (66.i) gives 



Lpp = 24' 



18.7 

 20.7 



(17,300)"' = 506.5 ft. 



Using for V the trial speed of 20.5 kt, 

 ^20.5 



Lpp = 24( ~| ) (17,300)''' = 515.3 ft. 



Baker's formula is put into nearly dimensionless 

 form by substituting V for A, and changing the 

 numerical coefficient accordingly. However, 

 Baker's definition for V remains somewhat 

 indefinite, and the ratio of Lpp to L^wl depends 

 upon the type of stern. Using a specific volume of 

 34.977 ft' per long ton for salt water, (Eq. 66.i) 

 becomes 



^iYTvf 



(34.977)''' 



Lpp = 7.33881 



V 



¥' 



(66. ii) 



2+ VJ 



For the trial speed of 20.5 kt, Eq. (66. ii) gives 

 20.5^ 



Lpp = 7.33881 p^ I (605,500)''' = 515.4 ft 



A sUght discrepancy in length is expected here 

 because the volume of 605,500 ft' was calculated 

 by assuming a specific volume of 35 ft' per ton 

 instead of the standard figure of 34.977 ft' per ton. 



If all the governing factors could be known, 

 properly weighed, and taken into account there 

 would undoubtedly be found a most efficient 

 length for each such set of conditions. Many 



designers are hesitant about exceeding that 

 length, even though they do not know what it is. 

 However, the excellent performance of many 

 ships in the past, after a lengthening process 

 which involved an increase of from 11 or less to 

 30 or more per cent of their original waterline 

 length, is an indication that too long a hull is by 

 no means the handicap that has been anticipated 

 in the past [Mar. Eng., 7 Jul 1954, pp. 66-67, 81]. 



A preliminary study of the wavegoing situation, 

 elaborated upon in Part 6 of Vol. Ill, indicates 

 that the greatest speed reduction is to be expected 

 when the ratio of wave length L^. to ship length 

 LnrL is from 0.8 to 1.0. Also, a study of available 

 data such as those in H.O. 602, 1947, reveals that 

 the maximum wave lengths to be expected in 

 the ocean areas traversed by the ABC ship are 

 of the order of 385 ft. For a ship length of 515 ft, 

 the ratio Lw/L^l is 385/515 or 0.747. This is 

 rather close for comfort to the low limit of 0.8, 

 but at least it does not he within the range 0.8 

 to 1.0. 



A somewhat different line of attack on the 

 length problem, still empirical and admittedly 

 taking account of quiet-water performance only, 

 is based upon data collected from the following 

 sources, among others: 



(1) Bates, J. L., Shipbuilding Encyclopedia, 1920, p. 200 



(2) Liddell, E., NECI, 19.34-1935, Vol. 51, pp. D45-D46 



(3) Nevitt, C. R., SNAME, 1945, Fig. 2, p. 316 



(4) Van Lammeren, W. P. A., RPSS, 1948, Fig. 39a, p. 89 



(5) Thayer, E., SNAME, 1948, Fig. 29, p. 409 



(6) Vincent, S. A., unpubl. Itrs. to HES, Sep 1947, Oct 



1952 



(7) SNAME Resistance Data sheets. 



These data, for merchant and combatant vessels 

 of orthodox form which have given good perform- 

 ance, cover a wide range of Taylor quotient, 

 fatness ratio or displacement-length quotient, and 

 longitudinal prismatic coeflicient Cp . They have 

 been checked and supplemented by comparison 

 with the proportions of models fisted on the 

 SNAME RD sheets which have bettered Taylor 

 Standard Series performance at and near the 

 designed speeds. The result is two pairs of 

 empirical curves on Fig. 66. A which bound two 

 design lanes. 



The upper pair defines the limits of displace- 

 ment-length quotient A/(0.010L)' and 0-diml 

 fatness ratio F/(0.10L)' on a base of T, and F„ 

 for good practice and normal designs. The 

 lower pair defines the limits of Cp in the same way. 

 However, the values for good designs may well 



