quiet water in the slip during stormy weather. Although some resonance of 

 the slip was expected, the model showed the slip was highly resonant to 

 wave periods typical of Pacific storms, and only a watertight door (or lock 

 gate) would provide quiet water in the slip. An alternate solution of 

 providing short stub breakwaters at each of the eastern corners of the 

 island was so effective in maintaining quiet water during most wave condi- 

 tions that the slip was replaced by a small wharf. In general, the other 

 features of the island plan were found to be satisfactory, as was the con- 

 cept of a high seaward face sheltering the lower-elevation island work 

 area. 



Using the same three-dimensional model, the feasibility of using two 

 concrete ship hulls as a separate, submerged breakwater seaward of the 

 island was investigated. By considering wave runup on the island's sides, 

 the effectiveness of different locations and spacing for the hulls was 

 studied. Although cost studies indicated possible savings in construction 

 cost, Richfield elected not to use the hulls because of the less attractive 

 appearance and possible adverse public reaction. 



The second series of laboratory tests involved two-dimensional wave 

 channel tests to evaluate the proposed seaward face revetment section. 

 The first test runs, with a wave height slightly below the design wave 

 height of 27 feet, verified stability for the design wave. Following 

 these initial tests, wave heights were increased by steps to a maximum 

 prototype height of 34 feet. 



As anticipated, the section showed dimage under attack of waves larger 

 than the design wave. A gratifying feature, however, was that the section 

 showed no tendency towards catastrophic failure due to any single wave, 

 but only gradually increasing damage. This was consistent with the basic 

 design objective of an economical section which might sustain damage, but 

 which would not endanger the whole island when subjected to rare large waves. 



The model tests were useful in developing the final shape of the 

 island and in verifying the west face revetment design and attenuation of 

 westerly waves. Prototype performance to date has been as predicted by the 

 tests . 



Testing was conducted in February 1956 to meet a 16 March 1956, dead- 

 line for incorporation in the design, which went to bid shortly thereafter 

 for a subsequent 25 May 1956, contract date. An optimum schedule would 

 provide for three-dimensional model testing at three stages of design pro- 

 gress. Design concepts should first be tested with a small-scale model 

 (about 1:150,000) allowing many tests at a reasonably low cost per test. 

 The objective would be to investigate many design concepts in a qualitative 

 manner. Next, as a given design is being developed, a major test program 

 should be conducted, using a large-scale model (1:70,000 or larger) for 

 definitive and quantitative input to the design development. As the design 

 stage nears completion some additional testing should be done to verify the 

 expected performance of the near-final design. Tne second and final stage 



24 



