Evaluation of Motions of Marine Craft in Irregular Seas 



DISCUSSION 



M. Fancev 



Institut Za Brodska Hidrodinamika 



Zagreb, Yugoslavia 



In dealing with any real problem of the ship motion stabilization one re- 

 ceives little help from a statistical treatment of sea and motion properties, but 

 rather has to look for a deterministic way of analysis. During my stay at 

 Davidson Laboratory in 1961, 1 tried first the Fourier Series synthesis with 

 known transfer functions and wave time history [13]. Although the results were 

 good, the impulsive response technique appeared more practical and simple in 

 application, and motion predictions (reproduced here in Fig. 4) proved to be re- 

 liable. Further extensive work at the Davidson Laboratory on this subject, as 

 may be seen from the paper, confirmed the validity and practibility of the im- 

 pulsive response technique. 



The authors are to be congratulated on the pedantic and precise development 

 of the theory for the covered modes of oscillations leaving so little to the intui- 

 tive way of reasoning. 



Wasn't this pedantry a little vague in the section following Eq. (42), as re- 

 gards the oddness and evenness of phase angle and amplitude of response, re- 

 spectively? These conclusions follow from a pure mathematical expression, 

 but a physical interpretation introducing negative frequencies could be easier 

 followed. 



The chapter on the shift of wave measurements cleared entirely the concep- 

 tion of physical realizability, which has to be understood in this specific applica- 

 tion of impulsive response technique. 



In the motion stabilization application some "free" future time will have to 

 be on disposal for the selection of proper orders for the control system of sta- 

 bilizers. One way to get this time "fund" is the measurement of waves well off 

 the ship, but this could be either impractical or unreliable. Practically, the 

 bow, the stern, and the sides of the ship are mostly too remote from C.G. to be 

 used for wave measurements. In such a case impulsive response will exhibit 

 values other than zero for "future" time, but because of the loss of the precision 

 of prediction, one could cut off a piece of the impulsive response to get the re- 

 quired "future" time. Here again the statistics could enter and make this trun- 

 cation in a proper way thereby giving the limitations and expectations of such a 

 procedure. 



Such a truncation of impulsive response (without any special statistical 

 treatment) I did in Ref. 13 and I succeeded in forecasting the pitch and heave 

 3.6 seconds in advance (full scale) with quite the same degree of precision as 

 without truncation (Figs. 14 and 15 of Ref. 13). It has to be emphasized that ir- 

 regular waves, in this case, were measured at the wave probe a half model 

 length ahead of the model. 



499 



