Understanding and Prediction of Ship Motions 



Some important results have been obtained in this effort. However, my presen- 

 tation is not very closely related to these efforts, for I shall discuss progress 

 toward what I call a "scientific solution" of the problems of ship motions. 



Perhaps I should be more specific in defining a "scientific solution." By 

 this 1 mean that one starts with a mathematical model of the fluid. It may be — 

 and in fact must be — a highly idealized model, but the implications of the ideali- 

 zation are probably well-understood in a general sense. To this mathematical 

 model, one must add a set of boundary conditions and also possibly initial con- 

 ditions, all of which should be stated as precisely and accurately as possible. 

 Even though the fluid is represented by an idealized model, the resulting prob- 

 lem is always intractable. Therefore one must put forth a set of additional as- 

 sumptions which reduces the problem to manageable proportions. When this 

 analytical problem has been solved, one makes calculations and compares them 

 with experimental data. There will be discrepancies, and so one goes all the 

 way back to the beginning and tries to relax one of the restrictive assumptions, 

 find a more general solution, etc., etc. 



Two parts of this process qualify it as a "scientific solution" by my defini- 

 tion, viz., all of the assumptions are stated at the beginning, and improvements 

 are made by modifying the assumptions rather than by trying empirically to 

 patch up faulty results. 



In practice, the engineer may not have the time to do all of this, or it may 

 be simply impossible. Still, he must make predictions. So, if he is a good engi- 

 neer, he improves his first poor predictions in any way he sees fit. This prog- 

 ress requires great ingenuity and skill, and its accomplishment is an essential 

 element in the working of our technocracy. However, I shall not discuss such 

 attempts, important though they may be. Other speakers here are much better 

 qualified for this, and I leave it to them. 



Summary of Contents 



Generally speaking, we wish ultimately to supply certain statistical infor- 

 mation to the ship designer. We may justify such an approach either by reason- 

 ing that he cannot really use more precise information or by accepting the fact 

 that we cannot hope to provide anything better. In either case, we begin with a 

 statistical description of the sea, assuming that the water motion can be de- 

 scribed as the sum of many simple sinusoidal waves, each of which is described 

 separately by the classical Airy formulas of linearized water wave theory. It 

 was the great contribution of St. Denis and Pierson (1953) to suggest (a) that the 

 statistical nature of the sea could be expressed by allowing the phases of these 

 components to take on random values and (b) that the response of a ship to the 

 sea was the sum of its responses to the various components. They only sug- 

 gested these hypotheses, and it may be claimed that both had been made earlier, 

 but these authors were the first to state them in precise, quantitative terms. 

 Their suggestion (a) relates more to the oceanographer's problem, and so I shall 

 not consider it here. However, (b) will be discussed in some detail, for it has 

 received much attention in recent years and it is at the heart of our problem. 



