210 



THEORY OF SEAKEEPING 



7 Concluding Remarks— General 



Chapters 2 and 3 have been written with these aims: 



1 To provide naval architects with tools for de- 

 signing more seakindly ships. 



2 To provide data on loads imposed on a ship's 

 structure. 



Ship motions in irregular seas are very complex and 

 cannot be understood on the basis of \dsual observation 

 alone. Designers of seakindly ships have attained 

 considerable success by their intuitive ability alone. 

 Nevertheless, this is an uncertain process as is manifested 

 by the widely different cjualities of ships in service. 

 While a designer's intuition cannot be replaced, it should 

 be exercised within the framework of hard facts. Intui- 

 tion which is not guided by facts degenerates into super- 

 stition. 



Thus far observations at sea have provided only 

 qualitative material. '^ The necessary ciuantitative 

 knowledge has been obtained from model tests and 

 theoretical considerations. Sea conditions can be sim- 

 plified in model testing and certain sea and .ship char- 

 acteristics can be segregated for more detailed in- 

 vestigation. However, even in a model, the dynamics 

 of motions is complex and cannot be understood fully. 

 As a consequence it becomes necessary to resort to theo- 

 retical considerations. Such successful investigators as 

 W. Froude, for example, employed all three methods; 

 sea observations, model testing, and theoretical con- 

 siderations. 



A sufficiently developed theory indicates an inexor- 

 able logic of events: given a certain set of conditions, 

 certain results must inevitably follow. In the field of 

 naval architecture two sets of conditions are present: 



1 The state of the sea. 



2 A multitude of characteristics resulting from ship 

 form and mass distribution. 



The first .set of conditions is not under a designer's 

 control, but he must have information about it. This 

 was the subject of Chapter 1. The second set of con- 

 ditions is under a designer's control in that he can choose 

 a ship's form and can control its mass distributif)n to a 

 limited extent. Once these are selected, a certain ship 

 behavior will result. The relationship between a .ship's 

 form as laid out f)n a drawing board and its hj^drody- 

 namic characteristics is not obvious. However, it can be 

 understood in the light of a suitable theory. Once the 

 ciuantitative relationship between the geometrical form 

 and its hydrodynamic and dynamic characteristics is 

 established, the designer must use his intuition in select- 

 ing an advantageous and practical combination of \-arious 

 features. 



Ship-motion theory has been developing slowly partly 

 because it involves both hydrf)dynamics and the dy- 

 namics of rigid bodies. Developments during the past 

 three decades of d_ynamics theory in the field of aero- 

 nautical engineering can be applied to ships. The glaring 



3* With the exci'|)ti()ii of the ])iii>t jiniject of Cartwright and 

 Rydill (1957). 



need at the moment is for further ilevelopnient of hy- 

 drodynamic theory. 



The foregoing remarks explain what may appear to be 

 a tendency toward the theoretical considerations in 

 Chapters 2 and 3. These remarks may be needed in 

 reference to the first objective; namel.v, the seakind- 

 liness of ships. A]5plication of ship-motion theory to 

 the .second objecti\'e (inlVtrmation on structural loading) 

 requires no apology. Here definite ciuantitative in- 

 formation is evidently required. Generally .speaking, 

 this information can be obtained only by the co-ordinated 

 u.se of theory, towing-tank experiments and ob.serva- 

 tions at sea. 



7.1 Applications of Ship Motion Theory. The 



"theory" invariably takes the form of a set of coupled 

 differential eciuations, the coefficients of which depend 

 on a ship's form and on its mass distribution. In the 

 simplest realistic form of ship motion (coupled heaving 

 and pitching in regular waves), there are two ec|uations 

 containing a total of twel\-e coefficients on the left- 

 hand sides. The effects produced by variations of .so 

 many (|uantities are scarcely self-evident. It also ap- 

 pears impossible to e\'aluate these effects \>y model 

 tests because each change of ship form affects simulta- 

 neously and in a different manner several coefficients. 

 In theoretical calculations it is possible to \'ary each 

 coefficient separately in order to investigate its effects. 

 This reiiuires making a .systematic series of calculations 

 similar in nature to the systematic series of model 

 tests. 



Until recently the theory of siiip moti(jns was too crude 

 and neglected too many factors to warrant such a series 

 of calculations. In the past few years, however, it has 

 been improved greatly. In the ca.se of hea\-ing and pitch- 

 ing motions in head .seas it has reached a practical degree 

 of de\'clopment. Up to now the work has concentrated 

 entirely on development of the theory and on demon- 

 stration of its \-alidity. Today a designer can use it in 

 order to predict the seakindliness of a new \es.sel in head 

 .seas in comparison with a prototype of known perform- 

 ance. For research intended to give a broader imder- 

 standing of ship motions, the series of calculations men- 

 tioned earlier is needed. 



The manual calculations which have been u.sed in the 

 development of the ship-motion theory are awkward and 

 time-consuming. It is recommended, therefore, that a 

 suitable programming of the calculations be made for 

 available digital computing machines. It is suggested, 

 howe\-cr, that excessi\-e automatism be guarded against. 

 The computational procedure should be broken up into 

 steps of moderate sizes and the results of each step should 

 be inspected. It should be kept in mind that the theory 

 is still new and developing. Therefore, it should be 

 possible to make changes in different parts of the pro- 

 gramming without upsetting the entire process. In 

 particular, it is suggested that the computation of co- 

 efficients be made a separate preliminary stej) prior to 

 soh'ing the eiiuations. By a judicious breakdown of 

 the entire programming into several steps, it should be 



