458 



HYDRODYNAMICS IN SHIP DESIGN 



Sec. 65.6 



as it were, because then the design problems are 

 visualized more clearly and they may be solved 

 in the most simple and direct manner. 



Using an alternative procedure, it is often 

 possible to work up a new hull design as a variant 

 or as a development of a previous design which 

 has proved itself in the same service and which 

 may be taken as a sort of parent form. This 

 procedure is facilitated by reference to the rather 

 comprehensive data now in existence on the 

 behavior of a multitude of ships and their models, 

 such as the SNAME Resistance Data sheets 1 

 through 160. The quantitative information re- 

 quired to set up a design and to carry it through 

 the successive steps must necessarily be derived 

 from model and ship performance data which are 

 known to be accurate and reliable. When in 

 doubt, the data are to be used with caution or, if 

 possible, only after a check and double check. 



The parent-form-copying or development pro- 

 cedure manifestly forms an unsound or at least 

 an uncertain basis for the design of a ship to 

 meet totally new and unexpected requirements. 

 A certain amount of improvement results from 

 successive developments of a given parent design 

 but more real progress is often made by starting 

 out afresh. Remember that the ship being copied 

 or modified was designed not yesterday but 

 several years ago. Its designer would be the 

 first to admit that much has been learned since 

 then. If starting out today, even he would not 

 reproduce the design exactly. 



65.6 Limits for Wavegoing Conditions to be 

 Encountered. To develop intelligently a good 

 ship design for wavegoing requires the establish- 

 ment, if practicable, of some sort of limits for the 

 wavegoing conditions to be encountered, more 

 precise than those of Table 64. d. Will most of 

 the waves be shorter or longer than the ship, or 

 about the same length? Will they be long and 

 regular, or short, steep, and confused? This can 

 be done on the basis of: 



(a) Existing knowledge of weather and waves 

 along the specified route in the various seasons of 

 the year, based on some sort of statistical analysis 

 of extensive data 



(b) Selection from the statistical data of the 

 characteristics and pattern of the predominant 

 wave, or the features of the one which is Ukely 

 to prove the most troublesome 



(c) An arbitrary declaration that the ship shall 

 give its best performance in a specified type of 



sea, and a deliberate acceptance of whatever 

 performance is obtained in other types of seas. 



Obviously, the design problem is materially 

 simplified if the ship is to operate on only one 

 run and to travel only in certain areas, especially 

 if the sea conditions in those areas are reasonably 

 consistent and predictable. 



65.7 The Bracketing Design Technique. In 

 a pioneering design, for which the basic physical 

 phenomena are somewhat uncertain and the 

 results of past experience are Umited or non- 

 existent, it frequently becomes necessary for a 

 designer to reach far into an unknown and un- 

 trodden field. He must commit himself and others 

 to the acceptance of risks or the expenditure of 

 funds which under normal circumstances could 

 not be justified as good engineering. In this 

 predicament he desperately needs any kind of 

 assurance, no matter what its source or reliability. 

 Fortunately, there is one means by which a ship 

 designer can extricate himself from a situation 

 such as this — a method which has been found 

 useful and successful in many other lines of 

 endeavor. 



The essence of this method is to determine the 

 extreme hmits of position in the unknown field, 

 then to fix a position between these limits by any 

 simple convenient method which appears suitable. 

 The region in which the unknown solution is to 

 be found lies somewhere between the two limits 

 and is thus bracketed by them. The position of 

 the region with respect to the limits, in other 

 words the spotting of the region where the solu- 

 tion probably Ues, involves a process of arithmetic 

 averaging or of estimating by a sort of mean 

 proportional of the hmits. 



A design problem of this kind developed with 

 the shaping of the alternative arch-type or tunnel 

 stern for the ABC ship, described in Sec. 67.16. 

 It was considered most important that the maxi- 

 mum fore-and-aft slope of the roof of the tunnel 

 be as large as possible, yet not so large as to 

 result in irregular flow or separation along the 

 tunnel roof. There was ample full-scale evidence 

 that on large ships with twin skegs, tunnel slopes 

 of 8 to 9 deg were satisfactory, with the roof 

 submerged a moderate amount below the at-rest 

 waterhne. There was evidence on some special 

 models that a centerplane slope of 30 deg, at the 

 same degree of submergence but on a convex 

 body surface and not in a concave tunnel roof, 

 was free from separation. It seemed reasonable to 



