Page 360 - Practical Design Ships and Floating Structures
P. 360

335

        third-order model are considerably different. The result of the thii-order model (Case 2) agrees more
        closely  with  the  exact  solution  (Case  3).  The  reason  for  the  difference between  the  result  of
        second-order model (Case  1) and the exact solution (Case 3) is considered to be that the region of
        interest,  in  this  example,  is  comparatively wide  to  approximate the  response  surface using  the
        second-order model.
        Figures 5  to 7 show the obtained results for these three cases.  The result of the third-order model
        (Case 5) agrees closely with  the exact solution (Case 6). The second-order  model  has a practical
        accuracy in  estimating the  peak value;  however,  the response  surface  is  slightly deformed.  The
        optimum design point, according to these results, is about  2 =37 to 38 and  nhg =I 0 to 1 1.
                                                             15
                                                             13
                                                             11
                                                            P9
                                                            =,
                                                             s
                                                             3
                                                             1
                                                              15  20  25  30  35  40  45  50  55
                     nu.                      nH                        n,
         Figure 2: Hull steel weight (ton)   Figure 3: Hull steel weight (ton)   Figure 4  Hull steel weight (ton)
            (Case 1 : Second-order model)   (Case 2: Thirdarder model)   (Case 3: Exact solution)

        Next, to examine the characteristics of the response near the peak value, re-calculation is carried out
        for a smaller region (Le.,  30 I nvr I60 and  5 Ink 5 15).













                                                              .   .  .  .   .
                     n                        n,             -30   35   40   45   50   55   60
         Figure 5: Hull steerwei  t (ton)   Figure 6: Hull steel weight (ton)   Figure 7: Hull #&I   weight (ton)
            (Case 4: Secondarg model)   (Case 5: Third-order model)   (Case 6: Exact solution)

            Case 4: Second-order model with 12 experimental points
            Case 5: Third-order model with 20 experimental points
            Case 6: Exact solution
        3.1.3 Optimization considering the construction cost

        In the above section, optimization of the structure in the view of minimum weight was performed by
        changing the number of vertical stiffeners and horizontal girders. According to the results obtained,
        optimum value was  found to  be  the  combination  in  which  the  number of  vertical stiffeners and
        horizontal girders is considerably more than the actual bulkhead structure (e.g.,  one actual tanker with
   355   356   357   358   359   360   361   362   363   364   365