step 5 — design detailed 



cross section for each alternative 



96. This is the second highly subjective step in the proposed optimi- 

 zation procedure where coastal engineers should, for each pair of design cri- 

 teria, prepare a cross-section design with all the detailed features appropri- 

 ate for the site conditions and other constraints. Practical considerations 

 discussed in Part II of this report should be incorporated. All the special- 

 ized experience and intuition available should be applied in this step, but it 

 must be applied consistently to each alternative. It is critical that bias be 

 studiously avoided at this stage. An estimate of the construction cost for 

 each alternative detailed cross section should be prepared at the conclusion 

 of this step. 



Step 6 — estimate wave transmission ' 

 characteristics of each alternative 



97. An analytical procedure should be performed at this point to esti- 

 mate the wave transmission characteristics as a function of incident waves 

 H^(H-) for each alternative. The program MADSEM (Seelig 1980a and WES, CERC 

 1984a) is useful for this purpose. The program accounts for the relative size 

 and permeability of each layer of the breakwater cross section and the rela- 

 tive runup characteristics of the armor layer. Wave transmission by overtop- 

 ping (Equations 26 through 29) and permeation (Madsen and White 1976) is es- 

 timated. The program is not as well verified for concrete armor units as for 

 quarrystone, but it serves well at this stage for comparative purposes. A 

 range of incident wave conditions should be simulated to obtain a substantial 

 set of H. (H.) points, including several more severe than the design condi- 

 tion. The incident wave conditions need to correspond to height and period 

 combinations predicted for the site in Step 1 . Wave period is a sensitive 

 factor for wave transmission, as applied in the program MADSEN. An appropri- 

 ate wave period (such as the peak spectral period T ) must therefore be as- 

 sociated with each (significant) incident wave height, as suggested in Step 1. 

 Transmitted waves are not Rayleigh distributed, as discussed in Part IV and 

 Andrew and Smith (in preparation), but can be represented by a single height 

 such as the root mean square wave height Hj,^^ or H^, c« . MADSEN predicts 

 the H of waves transmitted by the combined effects of both permeation and 

 overtopping. 



64 



