180 



THEORY OF SEAKEEPING 



after the publication of a paper by St. Denis and Pierson 

 (19.53) on ship motions in irregular waves. At the same 

 time, the definitions of sea-wave energy spectrum by 

 Darbyshire and Neumann (.see Chapter 1) indicated that 

 irregular wa\-es cannot be taken as mere disorderliness 

 of wave motion, but must contain definite statistic'al 

 characteristics. Artificially produced irregular waves 

 must ha-\-e the type of energy spectrum found at sea. 



Working with a plunger type of wa\'e-maker, E. V. 

 Lewis (1955 and 19566) produced irregular waves liy 

 varying the potential supplied to the driving motor from 

 stroke to stroke b.y means of a mechanically driven step 

 switch. The connections of 25 alternate potentials were 

 arranged randomly and the resultant wa\'es were found 

 by analysis to resemble a partially developed Neumann 

 .spectrum (Lewis and Numata, 1956). Although the 

 electric potential was varied step-wise, by the time the 

 motor was accelerated or decelerated and by the time 

 the waves were generated, the spectrum became con- 

 tinuous. When a fixed stroke of the plunger is used, 

 the amplitudes of short waves become excessive. This 

 is corrected b,y the use of light plastic foam floats, so 

 proportioned as to act as filters attenuating short waves, 

 while having little effect on the long ones. In E. V. 

 Lewis' work, waves were produced of spectra similar to 

 typical irregular sea spectra. Fuchs (1956), on the 

 other hand, attempted to reproduce a specific wa\'e 

 pattern obtained from a sea record. 



The waves ju.st described have parallel crests normal 

 to the tank length, and are irregular only m the sense 

 of periods and amplitude variation. Such waves us- 

 ually are referred to as "long-crested" and approximate 

 the so-called regular swells in Nature, which in reality 

 are never regular. The normal wind-produced sea is 

 "short-crested" and is vi.sualized as composed of wave 

 components of many directions. The generation of 

 .such waves in rectangular or scjuare tanks now becomes 

 of interest. A brief summary (if this process of wave 

 generation was given by Marks (1956). 



4.4 The Nature of Activity in Towing Tank Testing 

 in Oblique Seas. In the foregoing paragraphs reference 

 was made to descriptions of eciuipment used for model 

 testmg in oblique and irregular wa\-es. These are now 

 becoming available for commercial use and research. 

 Literature on contemplated types of testing and on or- 

 ganization of research activities in these tanks however, 

 is alarmingly meager. The only publications in this field 

 appear to be those of Korvin-Kroukovsky and Lewis 

 (1955a) and E. V. Lewis (1956c?). Newton (1957) pre- 

 sented a formal discussion at the Eighth International 

 Towing Tank Conference. Discussions of prospecti\'e 

 testing activity are valuable in stimulating research and in 

 elimmating pitfalls which are encountered in model testing 

 in waves. Such discussions also will miprove interpreta- 

 tion of test data. In this connection a determined effort 

 should be made to advance the three-dimensional theory 

 of ship motions, i.e., in six or seven (with rudder) degrees 

 of freedom, in order to present errors and to assist in the 



interpretation of results. There are man}' possibilities 

 for \-aluable achievement. 



The activity of oblique-wave towing tanks can be 

 broadly divided into : 



1 Tests in which the model behavior is assumed to 

 represent directlj- the ship motions at sea. 



2 Tests made in connection with the development 

 and application of a .ship motion theory. 



4.41 Tests of practical nature; conditions of simi- 

 larity. This heading is u>('d for hrcxity to cox'er tests 

 in which the model beha\'ior is considered to represent 

 directly the .ship motions at sea. While this definition 

 eliminates use of theoretical calculations beyond the 

 usual reduction of test data, the theory of multimode 

 ship motions must be kept in mind for the proper design 

 of experiments and in the interpretation of the data. 

 Because of the current lack of a sufficiently developed 

 theor.y, "practical" tests probably will predominate in 

 towing-tank activity in the near future, as has been 

 emphasized liy Newton (1957). For want of a developed 

 theory also the following discussion must be limited to 

 qualitative and intuitive aspects. 



If the model behavior is to represent ship motions cor- 

 rectly, a complete similarity of ambient conditions must 

 be provided. Lack of a workable theory makes it im- 

 pos.sible to apply corrections for deviations of model 

 te.st conditions from true ship conditions. In addition 

 to the traditional conditions, the law of similarity applies 

 to the following factors (m growing order of importance) : 

 (a) ]\Iass distribution; (b) Rudder control; (c) Wave 

 structure. 



The proper scaling of weights and of moments of 

 inertia has always been obser^-ed in towing tanks. In 

 coimection with tests in oblique waves it becomes neces- 

 sary also to observe the proper orientation of the prm- 

 cipal axes of inertia. The major practical difficulty 

 here probably is the securing of .sufficiently detailed 

 information on the mass distribution in actual ships. 



The powerful yaw-roll coupling of ships makes it 

 nece.s.sary to observe similarity of rudder control used 

 to maintaui a model and a ship on course. In the 

 theoretical literature the rudder-control function is 

 usually described in the form 6 = 5 (J'\pdt, xp, 4/, \f), 

 in which a limited number of the functions is chosen. 

 It has been recommended (Dalman and ^'ossers, 1957) 

 that such a control be used in model tests. While this 

 may gi\'e a realistic control presentation for future 

 ships, it is not realistic for most contemporary ships. 

 It appears that the current practice at sea is to shift to 

 manual control even in .slightly rough seas because of 

 imperfecticjns of the automatic steering devices. It 

 becomes important, therefore, to investigate empirically 

 properties of commonly used manual steering and meth- 

 ods of reproducing it in a model. Conceivably this can 

 be accomplished by a suitable application of statistical 

 theory. Simultaneou.'ily obtained records of a ship's 

 yawing and rudder motions are needed as basic material 

 for such a study. 



It is evident that towing-tank waves must correspond 



